Digital Photography in the Diagnosis and Follow-up of ... · atrophy and serous detachment of the ciliary body with ocular hypotony together with the shallow retinal detachment of

Department of Ophthalmology

University of Helsinki

Helsinki, Finland

Digital Photography in the Diagnosis and Follow-up of Ocular Diseases

by

Jukka M. Saari

Academic Dissertation

To be publicly discussed, by permission of the Medical Faculty of the University of

Helsinki, in the Auditorium 2 of Biomedicum Helsinki, Haartmaninkatu 8, Helsinki on

November 30th, 2007 at 12 o’clock noon.

From the Department of Ophthalmology, University of Helsinki, Helsinki, Finland, and

the Department of Ophthalmology, University of Turku, Turku, Finland

Supervised by:

Professor Tero Kivelä, MD, FEBO Professor K. Matti Saari, MD, FEBO

Department of Ophthalmology Department of Ophthalmology

Helsinki University Central Hospital University of Turku

Helsinki, Finland Turku, Finland

Reviewed by:

Docent Eero Aarnisalo, MD Docent Markku Teräsvirta, MD, FEBO

Department of Ophthalmology Department of Ophthalmology

Satakunta Central Hospital Kuopio University Central Hospital

Pori, Finland Kuopio, Finland

Opponent:

Professor Hannu Uusitalo, MD

Department of Ophthalmology

University of Tampere

Tampere, Finland

Yliopistopaino 2007, Helsinki

ISBN 978-952-92-3034-1

ISBN 978-952-10-4374-1 (PDF version, available at http://ethesis.helsinki.fi)

To my parents

- 4 -

ABBREVIATIONS

AMD Age-related macular degeneration

CCD Charge-coupled device

CI Confidence interval

CNV Choroidal neovascularization

COMS Collaborative Ocular Melanoma Study

CT Computed tomography

DR Diabetic retinopathy

ETDRS Early Treatment Diabetic Retinopathy Study

FAG Fluorescein angiography

HLA Human leukocyte antigen

ICD International classification of diseases

ICG Indocyanine green

IOP Intraocular pressure

IR Infrared

IRMA Intraretinal microvascular abnormalities

IRT Infrared transillumination

LBD Largest basal diameter

MA Microaneurysm

MRI Magnetic resonance imaging

NPDR Non-proliferative diabetic retinopathy

OCT Optical coherence tomography

PAD Pathologic-anatomical diagnosis

PAL Phase alternating line

PC Personal computer

PDR Proliferative diabetic retinopathy

RNFL Retinal nerve fiber layer

RPE Retinal pigment epithelium

SLO Scanning laser ophthascope

TRC Topcon retinal camera

UV Ultraviolet

VKH Vogt-Koyanagi-Harada

WHO World Health Organization

- 5 -

LIST OF ORIGINAL PUBLICATIONS

This dissertation is based on the following original publications, which will be referred

to in the text by their Roman numerals I-V.

I Saari JM, Summanen P, Kivelä T, Saari KM. Sensitivity and specificity of

digital retinal images in grading diabetic retinopathy. Acta Ophthalmol Scand

2004; 82: 126-130.

II Saari JM, Kivelä T, Summanen P, Nummelin K, Saari KM. Digital imaging in

differential diagnosis of small choroidal melanoma. Graefes Arch Clin Exp

Ophthalmol. 2006; 244: 1581-1590.

III Saari JM, Nummelin K. Digital infrared transillumination imaging of iris. J

Ophthalmic Photogr 2005; 27(1): 20-24.

IV Saari JM, Nummelin K. Iris atrophy, serous detachment of the ciliary body and

ocular hypotony in chronic phase of Vogt-Koyanagi-Harada disease. Eur J

Ophthalmol 2005; 15: 277-283.

V Saari JM, Kivelä T, Summanen P, Nummelin K, Saari KM. Infrared

transillumination imaging and fluorescein angiography of iris nevus and

melanoma. J Ophthalmic Photogr 2007; 29(1): 17-20

The original publications have been reproduced with the permission of the copyright

holders.

- 6 -

TABLE OF CONTENTS

ABBREVIATONS ………………………………………………………………………… 4

LIST OF ORIGINAL PUBLICATIONS………………………………………………… 5

TABLE OF CONTENTS…………………………………………………………………..6

ABSTRACT………………………………………………………………………………... 9

1. INTRODUCTION……………………………………………………………………... 11

2. REVIEW OF LITERATURE……………………………………………………….... 13

2.1 Photographic and digital imaging methods in ophthalmology……………….. 13

2.1.1 Photography of the anterior segment of the eye................................................... 13

2.1.2 Slit-lamp photography.......................................................................................... 13

2.1.3 Transpupillary transillumination photography of the eye……………………… 14

2.1.4 Transcleral transillumination photography of the iris………………………….. 14

2.1.5 Infrared photography of the iris………………………………………………… 15

2.1.6 Infrared transillumination stereophotography of the ciliary body and iris……... 16

2.1.7 Photography of the ocular fundus………………………………………………. 17

2.1.8 Monochromatic fundus photography……………………………………………19

2.1.8.1 Red-free light photography………………………………………………….. 19

2.1.8.2 Red light photography………………………………………………………. 22

2.1.9 Infrared photography of the ocular fundus……………………………………... 22

2.1.10 Fluorescein angiography……………………………………………………….. 23

2.1.10.1 Fluorescein angiography of the ocular fundus……………………………… 25

2.1.10.2 Fluorescein angiography of the iris…………………………………………..25

2.1.10.3 Anterior segment fluorescein angiography in corneal diseases……………...26

2.1.11 Digital imaging of the eye……………………………………………………… 27

2.1.12 Indocyanine green angiography…………………………………………………28

2.1.13 Scanning laser ophthalmoscopy………………………………………………... 29

2.1.14 Optical coherence tomography…………………………………………………. 30

2.1.14.1 OCT retinal imaging………………………………………………………… 31

2.1.14.2 OCT anterior segment imaging……………………………………………… 32

2.2 Teleophthalmology………………………………………………………………. 33

2.3 Grading of diabetic retinopathy………………………………………………… 34

2.4 Melanocytic tumours of the uvea……………………………………………….. 37

2.4.1 Choroidal nevus and melanoma…………………………………………………38

2.4.2 Nevus and melanoma of the iris…………………………………………………39

2.5 Vogt-Koyanagi-Harada disease………………………………………………….40

3. AIMS OF THE PRESENT STUDY…………………………………………………...43

4. MATERIALS AND METHODS………………………………………………………45

4.1 Subjects (I – V) …………………………………………………………………...45

- 7 -

4.2 Diagnostic criteria (I, II, IV, V) ………………………………………………… 47

4.2.1 Diabetes (I) …………………………………………………………………….. 47

4.2.2 Grading of diabetic retinopathy (I) …………………………………………….. 47

4.2.3 Uveal melanoma (II, V) ………………………………………………………... 49

4.2.3.1 Choroidal melanoma (II) …………………………………………………….49

4.2.3.2 Iris melanoma (V) …………………………………………………………… 50

4.2.4 Uveal nevus (II, V) …………………………………………………………….. 50

4.2.4.1 Choroidal nevus (II) ………………………………………………………….50

4.2.4.2 Iris nevus (V) …………………………………………………………………51

4.2.5 Vogt-Koyanagi-Harad disease (IV) ……………………………………………. 51

4.3 Classification of iris colour (III, V) …………………………………………….. 53

4.4 Imaging methods (I – V) …………………………………………………………54

4.4.1 Digital retinal imaging (I, II, IV) ………………………………………………. 54

4.4.1.1 Digital retinal imaging in screening for DR (I) …………………………….. 54

4.4.1.2 Digital retinal imaging of choroidal melanocytic tumours (I) ……………… 55

4.4.1.3 Digital retinal imaging of the Vogt-Koyanagi-Harada patient (IV) ……….. 55

4.4.2 Colour photography of the anterior eye (III, IV, V) …………………………… 55

4.4.3 Digital transpupillary transillumination imaging of the iris (IV) ……………… 55

4.4.4 Infrared transillumination imaging of the iris and ciliary body (III, IV, V) ……56

4.4.4.1 Infrared transillumination stereophotography of the iris and ciliary

body (V) ………………………………………………………………………56

4.4.4.2 Digital infrared transillumination imaging of the iris and cialiary body

(III, IV, V) …………………………………………………………………… 56

4.4.5 Fluorescein angiography (II, IV, V) …………………………………………… 57

4.4.5.1 Fluorescein angiography of the fundus (II, IV) ……………………………... 57

4.4.5.2 Fluorescein angiography of the iris (V) …………………………………….. 59

4.5 Subtraction method to evaluate growth of choroidal melanocytic tumours

(II) ………………………………………………………………………………… 59

4.6 Data collection, enrolment and experimental protocol (I, II, V)……………… 60

4.6.1 Experimental protocol and grading of diabetic retinopathy (I) ………………... 60

4.6.2 Eligibility criteria, enrolment and experimental protocol of small choroidal

melanoma study (II) …………………………………………………………… 60

4.6.3 Evaluation of the usefulness of the subtraction method (II) …………………… 62

4.6.4 Data analysis of IRT and FAG findings of iris melanocytic tumours (V) …….. 62

4.7 Statistical methods (I, II, III, V) ………………………………………………... 63

5. RESULTS………………………………………………………………………………. 65

5.1 Sensitivity and specificity of digital retinal images in grading diabetic

retinopathy (I)……………………………………………………………………. 65

5.2 Digital imaging in differential diagnosis of small choroidal melanoma (II)…. 67

5.3 Digital infrared transillumination imaging of normal iris (III)………………. 72

- 8 -

5.4 Iris atrophy, serous detachment of the ciliary body and ocular hypotony

in chronic phase of VKH disease (IV)…………………………………………... 73

5.5 Infrared transillumination imaging and fluorescein angiography of iris

nevus and melanoma (V)………………………………………………………… 75

5.5.1 Infrared transillumination imaging)……………………………………………. 75

5.5.2 Fluorescein angiographic findings)…………………………………………….. 76

6. DISCUSSION)………………………………………………………………………..... 77

6.1 Sensitivity and speficifity of digital retinal images in grading diabetic

retinopathy (I)……………………………………………………………………. 77

6.1.1 Digital red-free and colour imaging in grading diabetic retinopathy (I) …………. 77

6.1.2 The effect of the size of the imaging field in grading diabetic retinopathy (I) …... 79

6.1.3 The role of screeners in grading diabetic retinopathy…………………………….. 79

6.1.4 Screening of diabetic retinopathy…………………………………………………. 79

6.2 Digital imaging in differential diagnosis of small choroidal melanoma (II)…. 80

6.2.1 Comparative use of digital colour, red-free light and red light imaging………….. 80

6.2.2 Subtraction method………………………………………………………………...82

6.2.3 Epidemiological aspects of choroidal melanoma………………………………..... 82

6.3 Digital infrared transillumination imaging of the iris (III, IV, V)…………..... 83

6.4 Infrared transillumination findings of the iris and ciliary body in

chronic phase of Vogt-Koyanagi Harada disease (IV)………………………… 90

6.5 Infrared transillumination imaging and fluorescein angiography in

differential diagnosis of iris nevus and melanoma (V) ………………………... 93

6.6 Teleophthalmology and future directions ……………………………………... 95

7. SUMMARY AND CONCLUSIONS…………………………………………………. 97

8. ACKNOWLEDGEMENTS…………………………………………………………… 101

9. REFERENCES………………………………………………………………………… 103

- 9 -

ABSTRACT

Purpose: The aim of the present study was to develop and test new digital imaging

equipment and methods for diagnosis and follow-up of ocular diseases.

Methods: The whole material comprised 398 subjects (469 examined eyes), including

241 patients with melanocytic choroidal tumours, 56 patients with melanocytic iris

tumours, 42 patients with diabetes, a 52-year old patient with chronic phase of VKH

disease, a 30-year old patient with an old blunt eye injury, and 57 normal healthy

subjects. Digital 50° (Topcon TRC 50 IA) and 45° (Canon CR6-45NM) fundus

cameras, a new handheld digital colour videocamera for eye examinations (MediTell), a

new subtraction method using the Topcon Image Net Program (Topcon corporation,

Tokyo, Japan), a new method for digital IRT imaging of the iris we developed, and

Zeiss photoslitlamp with a digital camera body were used for digital imaging.

Results: Digital 50° red-free imaging had a sensitivity of 97.7% and two-field 45° and

50° colour imaging a sensitivity of 88.9-94%. The specificity of the digital 45°-50°

imaging modalities was 98.9-100% versus the reference standard and ungradeable

images that were 1.2-1.6%. By using the handheld digital colour video camera only, the

optic disc and central fundus located inside 20° from the fovea could be recorded with a

sensitivity of 6.9% for detection of at least mild NPDR when compared with the

reference standard.

Comparative use of digital colour, red-free, and red light imaging showed 85.7%

sensitivity, 99% specificity, and 98.2 % exact agreement versus the reference standard

in differentiation of small choroidal melanoma from pseudomelanoma. The new

subtraction method showed growth in four of 94 melanocytic tumours (4.3%) during a

mean ±SD follow-up of 23 ± 11 months.

The new digital IRT imaging of the iris showed the sphincter muscle and radial

contraction folds of Schwalbe in the pupillary zone and radial structural folds of

Schwalbe and circular contraction furrows in the ciliary zone of the iris.

The 52-year-old patient with a chronic phase of VKH disease showed extensive

atrophy and occasional pigment clumps in the iris stroma, detachment of the ciliary

body with severe ocular hypotony, and shallow retinal detachment of the posterior pole

in both eyes.

- 10 -

Infrared transillumination imaging and fluorescein angiographic findings of the iris

showed that IR translucence (p=0.53), complete masking of fluorescence (p=0.69),

presence of disorganized vessels (p=0.32), and fluorescein leakage (p=1.0) at the site of

the lesion did not differentiate an iris nevus from a melanoma.

Conclusions: Digital 50° red-free and two-field 50° or 45° colour imaging were

suitable for DR screening, whereas the handheld digital video camera did not fulfill the

needs of DR screening. Comparative use of digital colour, red-free and red light

imaging was a suitable method in the differentiation of small choroidal melanoma from

different pseudomelanomas. The subtraction method may reveal early growth of the

melanocytic choroidal tumours.

Digital IRT imaging may be used to study changes of the stroma and posterior

surface of the iris in various diseases of the uvea. It contributed to the revealment of iris

atrophy and serous detachment of the ciliary body with ocular hypotony together with

the shallow retinal detachment of the posterior pole as new findings of the chronic

phase of VKH disease. Infrared translucence and angiographic findings are useful in

differential diagnosis of melanocytic iris tumours, but they cannot be used to determine

if the lesion is benign or malignant.

Key words: choroid, diabetic retinopathy, digital imaging, infrared transillumination,

melanoma, nevus, ocular fundus, photography, Vogt-Koyanagi-Harada disease

Introduction

- 11 -

1. INTRODUCTION

A photographic image requires light reflected from a target. The eye is in the very

surface of the body and the anterior segment of the eye is thus ready for photographic

imaging. The eye itself is a physiological “camera” of the body with a clear optic media.

Since the invention of the ophthalmoscope by Helmholtz in 1851, the unique possibility

of studying the ocular fundus in vivo has existed.101 The fundus can be photographed,

because it reflects a significant fraction of the light incident on it. The reflected light

forms an exact real image of the fundus, which can be documented on photographs. By

using monochromatic light of different wavelengths, different layers of the retina can be

photographed. The IR light penetrates the sclera and the iris, and it can be used for IR

transillumination photography of the iris and ciliary body.

Ophthalmic photography is an important part of the documentation of different

pathological conditions of the eye. The advantages of it include reliable detection and

accurate recording of different lesions and the possibility to compare the changes with

previous appearances. The colour slides and film-based photographs, however, are

difficult to manipulate, store, and transmit across telephone lines to other clinicians.

They are relatively expensive and evaluation of these images is delayed by the film

development.

The development of digital imaging technology during the course of the last ten years

has made it possible to switch to ophthalmic digital imaging. The advantages of it are

immediate observation of the recordings, replacement of a faulty image with a new one,

immediate assessment for treatment, an enlarged image of the lesion on a computer

Introduction

- 12 -

monitor screen, no expenses for film developing, convenient storage of the images, easy

comparison of follow-up pictures, and faster data transfer through telemedicine. The

present investigation was initiated to develop and evaluate digital imaging methods for

the examination of eye diseases.

Review of Literature

- 13 -

2. REVIEW OF LITERATURE

2.1 Photographic and digital imaging methods in ophthalmology

2.1.1 Photography of the anterior segment of the eye

The anterior segment of the eye is readily accessible for minute and delicate

examination. Straightforward photography of the external eye is useful in maintaining a

permanent record and in the assessment of small changes of a lesion. Photography of

the anterior segment of the eye was introduced by Drüner in 1900.56 Stereoscopic

cameras were brought into use by Lentz in 1924 and Drüner in 1927.57,139 Colour

photography of the anterior eye was started by Morax in 1929 and Redway in

1930.160,179 A useful instrument for the photography of the anterior segment of the eye

consisted of a simple camera mounted on standard ophthalmic equipment with a chin

and forehead-rest to immobilize the patient’s head.

2.1.2 Slit-lamp photography

Gullstrand developed the slit-lamp in 1911 and combined it with a corneal

microscope.93,94 In this system focal illumination of extreme accuracy and mobility was

combined with a binocular microscope giving an erect image with a full stereoscopic

effect. Goldmann added a joystick control for fine simultaneous adjustments of the slit-

lamp and the microscope.92 The system provided magnifications approaching those of

histology and therefore the technique was termed biomicroscopy.

The first photographic recordings of slit-lamp views were restricted to one plane and

were often not satisfactory.230 Photography of the anterior segment requires a camera

with a good depth of focus. Sysi incorporated a minute camera in one objective of the


- 14 -

biomicroscope.226 This allowed a high degree of accuracy in focusing. By using a

complicated apparatus Goldmann obtained pictures of a sharp optical section through

the cornea and the lens.91 Littmann’s anterior segment camera had a focal depth of

1:46.142

2.1.3 Transpupillary transillumination photography of the eye

If white light enters the eye in a parallel beam, the pupil appears luminous due to the

emerging rays of the fundus glow and it reveals lenticular opacities as well as atrophies,

and other defects of the pigment epithelium of the iris.55,72 Transpupillary

transillumination photographs may be taken by employing a camera with its focus at the

lens plane.144

2.1.4 Transcleral transillumination photography of the iris

Donaldson lead a bright light from a 600 watt-second electronic flash into the eye

through the conjunctiva and sclera by using a light pipe made of Pyrex glass. For

photographing iris translucency he used a stereoscopic anterior-segment camera and

colour film (Kodachrome II).53 Defects in the pigment epithelium showing translucency

of the iris were reported in albinism, diabetes mellitus, pigmentary glaucoma, the use of

miotics in open-angle glaucoma, pseudoexfoliation, and uveitis.54

For iris translucency photography Repo et al. lead the light from a diaprojector and

flashlight into the eye by using optical fiber with two branches at its proximal root. The

probe of the optical fiber was placed on the temporal portion of the eye as posteriorly as

possible. The photographs were recorded by using an anterior segment camera

(Topcon), attached to a biomicroscope, using black-and-white film (ASA 400).181,182,183


- 15 -

They reported peripheral iris transluminance in 45% of the patients with

pseudoexfoliation syndrome and capsular glaucoma and in 18% of the eyes in the

control group.183 They observed generalised transluminance of the iris in 42-45% of the

eyes of 62 unselected patients with carotid transient ischemic attacks.182 Abnormal iris

transluminance occurred also in patients with type 2 diabetes after 10 years of follow-

up, especially in those patients with severe retinopathy.239

2.1.5 Infrared photography of the iris

The range of wavelengths of the visible spectrum is from 400 nm at the blue end to 700

nm at the red end.35,40 The human eye is markedly less sensitive to light above 700 nm

wavelength, but intense light can be detected up to 780 nm, and will be perceived as red

light. The wavelength of the IR radiation is between 780 nm and 1 mm. Infrared

radiation cannot be seen like visible light. In ophthalmological literature the term IR

“light” is commonly used instead of IR radiation and we do so in this study as well. In

the same way we use here IR “transillumination” instead of IR transradiation. The

subdivisions of the IR region can vary depending on the publication. The International

Commission on Illumination (CIE) divides the IR spectrum into IR-A (780 - 1400 nm),

IR-B (1400 – 3000 nm) and IR-C (3 µm – 1 mm). IR-A is often called near-IR and IR-B

short-wavelength IR. The IR-C section has been divided sometimes as mid-wavelength

IR (3 – 8 µm), long-wavelength IR (8 – 15 µm) and far IR (15 - 1000 µm). The area of

IR-A and IR-B is sometimes referred to as reflected IR, and IR-C as thermal IR. The IR

rays used in ophthalmic photography and digital imaging are near-IR waves, which a

human being cannot feel.


- 16 -

Robert W. Wood published in 1910 the first IR photograph taken on experimental film

that required very long exposure time. In 1933, Dekking first published a paper on the

use of IR light in ophthalmology.45 IR photographs were taken on black and white IR

film. The IR photographs could be taken in an illuminated room when the visible light

was filtered by using a Kodak Wratten 25A filter in front of the film.71 A black room

was needed for the IR photography if a Kodak Wratten 87, 88A, or 89A filter was fitted

in front of the illuminating light.71 Due to the difference in the wave length between the

visible light and the IR light, the focusing distance of IR photography was greater than

by using the visible light. Thus the focusing was more difficult in IR photography than

in typical photography. Usually IR photography only revealed the superficial layer of

the iris and not the sharp definition of the structures. Because IR light passes easily

through the cornea, sclera, and pigment epithelium, it could be used for evaluation of

the anterior surface of the iris, presence of synechiae and the contour and size of the

pupil through the opacified cornea.71,110,150

An infrared TV camera could be used for IR photography of the iris.152 Better focusing,

the possibility of following dynamic changes of the iris and pupil and bigger primary

enlargement were benefits of this method. The equipment, however, was expensive and

resulted in a rather low sensitivity.152

2.1.6 Infrared transillumination stereophotography of the ciliary body and iris

In 1976, Saari and Nieminen described an analog IRT stereo technique for studying

deeper layers of the iris and the ciliary body.196 The photographs showed that the iris

was rather dense in children, well developed in adults and somewhat atrophic in the

elderly. On the posterior surface of the iris the radial contraction folds and the structural


- 17 -

folds of Schwalbe and circular contraction folds were seen; they were delicate in

children and more pronounced in adults and the elderly.199 IRT stereophotography was

used to show atrophic areas, foreign bodies, and cysts of the iris and ciliary body.196 It

showed different types of atrophic changes of the iris in Fuchs’ heterochromic cyclitis,

rheumatic iridocyclitis, and in sarcoid uveitis.121,197,200

Wilson measured the lens diameter during accommodation in a young female carrier of

ocular albinism by using retroillumination IR video photography and pixel unit

measurements. 248 He disproved the theory that the lens diameter increases in

accommodation. Roberts succeeded by using a standard digital camera modified to

detect visible and IR light, and a halogen fiberoptic light source placed against the

inferior eyelid, to demonstrate the presence of bilateral iris and ciliary body cysts in a

50-year old woman.185

2.1.7 Photography of the ocular fundus

The ophthalmoscope was introduced by Helmholtz in 1851.101 After its introduction

many improvements have been incorporated into the ophthalmoscope and innumerable

such modifications have been made by ophthalmologists and instrument makers.

Dennett introduced the electric ophthalmoscope in 1885.49 Gullstrand was the first to

construct the reflexless indirect ophthalmoscope in 1911.93

The first photograph of the human retina was taken by Jackman and Webster in 1886.

However, Nordenson in 1925 was the pioneer to develop successful photography of the

ocular fundus.165 The Nordenson camera (Zeiss) was a modification of Gullstrands

reflexless indirect ophthalmoscope, and all subsequent instruments are based on the


- 18 -

same general principles. The use of electronic and xenon arc flashes resulted in bright

illumination of the ocular fundus during photography.12,95 In 1955 Littman introduced a

new Zeiss fundus camera which was in general use for the next 20-25 years until the

new Japanese wide-angle cameras came into common use.143

Fundus cameras are often described by the angle of view. An angle of 30 degrees,

considered the normal angle of view, creates a film image magnified by a factor of 2.5.

A narrow-angle fundus camera has an angle of view of 20 degrees or less. Wide-angle

fundus cameras were introduced in the 1970s. They image between 45 and 140 degrees

and provide proportionately less retinal magnification.232

Both mydriatic and nonmydriatic fundus cameras require pupil dilation. Usually fundus

photography is done after pharmacologic pupil dilatation. Nonmydriatic fundus cameras

are used in a darkened room providing the natural pupil dilation. An IR light is then

used to preview the retina on a video monitor. Once the monitor’s image is focused and

aligned, a flash is fired, and the image is exposed. The nonmydriatic fundus camera was

pioneered by Canon in the late 1970s, and by the mid-1980s, Canon, Kowa, and Topcon

were marketing various models. Stereo fundus photography or any type of ocular

angiography cannot be performed with nonmydriatic cameras, as these techniques

require multiple flashes.232

Stereoscopic photography of the fundus is an old invention.153,231 Stereoscopy could

easily be attained if the camera was moved appropriately between exposures or by using

a stereo-camera fitted with appropriate prismatic devices.27,135,149,153,231


- 19 -

The early photographs were taken on monochrome black and white film. Mann, Von

Bahr, and Hansell and Beeson succeeded in colour photography of the eyeground by

using the Zeiss Nordenson fundus camera and Kodachrome colour film.12,95,149 Colour

photography was generally more informative than the black and white photographs

taken with white light. Photographs of the ocular fundus had a detailed exactness that

could not be equaled by any written word or free-hand reproduction.149

2.1.8 Monochromatic fundus photography

Monochromatic fundus photography is performed with the use of a narrow-band

spectral illumination of various wavelengths. Lesions on different anatomical layers in

the ocular fundus can be seen with an increased contrast when using appropriate

monochromatic illumination. The shorter waves of the spectrum allow visualization of

changes in the nerve fibre layer and sharp contours of the retinal vessels. Increasing the

wavelength of illumination increases the light penetration into the fundal layers.47,79

2.1.8.1 Red-free light photography

In 1913, Vogt was the first to observe the retina in red-free light.238 He gave a detailed

description of the RNFL. In 1956, Behrendt and Wilson introduced red-free light

photography of the ocular fundus.17 They used interference filters and black and white

film. They noticed that the RNFL was invisible in red light, but its visibility was

increased in green-blue and blue light from 549 to 477 nm though it began to disappear

at 431 nm. Blue light does not penetrate beyond the RNFL, but it is reflected from this

layer of the retina. With blue light the deeper retina is more or less invisible except for

locations where the RNFL is destroyed, which enhances the contrast between normal


- 20 -

and degenerated areas of RNFL.16 Several later authors confirmed the usefulness of

monochromatic red-free light photography in the visualization of the RNFL.47,79,157,234

Table 1 Percent transmittance of Kodak Wratten filters No 58 (green), 25 (red) and 87

(visibly opaque for infrared photography)62

Wavelength Percent transmittance____________________________________(nm) No 58 No 25 No 87

460 - - - 70 0.23 - - 80 1.38 - - 90 4.90 - - 500 17.7 - - 10 38.8 - - 20 52.2 - - 30 53.6 - - 40 47.6 - - 60 27.8 - - 80 9.0 - - 90 3.5 12.6 - 600 1.5 50.0 - 10 0.41 75.0 - 20 - 82.6 - 40 - 86.7 - 60 - 88.2 - 80 - 89.0 - 90 - 89.3 - 700 0.53 89.5 - 20 - 40 0.8 60 12.3 80 33.2 800 56.9 20 68.0 40 76.3 60 79.5 80 80.7 900 81.9 20 82.7 40 83.4 60 84.0 80 84.6 1000 85.3 40 86.6 80 88.1 1100 88.5

In 1972, Hoyt and Newman were the first to suggest that red-free light photography

might be useful for studying structural changes of RNFL in glaucoma in 1972.107 At


- 21 -

first the RNFL was photographed on black and white film usually through a Kodak

Wratten green filter No. 58 (540 nm, Table 1) using the 30° picture angle of a Zeiss

fundus camera.1 Later on a wide-angle fundus camera with its blue interference filter

(495 nm) could be used to photograph RNFL in glaucoma.2 The wide picture angle

allowed observation of the entire RNFL defect in a single picture. In older patients with

nuclear sclerosis the blue light was scattered in the anterior eye and better photographs

were achieved with longer wave-lengths (Kodak Wratten No 58, Table 1). Hemoglobin

transmits spectral red while it absorbs the green and blue-violet end of the spectrum.

Blood appears black in red-free light and the retinal vessels are sharply outlined against

a bright background.47,79,100

2.1.8.2 Red light photography

Red light photography, using a Kodak Wratten filter No 25 (Table 1) or 29, is useful for

examination of the deeper layers of the retina and choroid. Wavelengths above 580 nm

cause reduction in contrast of retinal vessels and a large increase in light penetration. It

shows the retinal arteries as light bands and the retinal veins dark.79 At 650 nm, the

largest arteries are barely visible, while the large veins can still be observed. At 600 –

650 nm it allows intense visualization of the choroidal pigment.47,79 In healing

toxoplasmic retinochoroiditis it revealed spotted accumulation of choroidal pigment at

the edges of the lesion, whereas only slight pigmentation was seen in colour

photographs.201 Red light photography revealed heavy pigmentation at the edges of an

old scar.202 With this method the choroidal vessels may be outlined as light bands

through the RPE. In choroideremia the red light photography clearly revealed the

choroidal vessels that were relatively obscured in red-free light photography.79


- 22 -

In juvenile haemorrhagic maculopathy red light photography revealed depigmentation

of the RPE overlaying the choroidal lesion, and clearly demonstrated the subsequent

pigment-ring lesion and proliferation of pigment epithelium at the borders of an old

lesion, RPE degeneration in the center of it, and pigment epithelial atrophy around the

lesion.194 Red light photography is valuable in showing pigmentary changes of the

earlier stages of AMD.

2.1.9 Infrared photography of the ocular fundus

Kugelberg photographed the ocular fundus with IR rays by using the Nordenson fundus

camera.136 The filter used was Kodak Wratten 25 (Table 1) and the plates were Eastman

IR plates 1-R. Kodak Wratten filter No 25 was used over the camera lens because IR-

sensitive film and plate materials are also blue and violet sensitive.62 The pictures

showed poor contrast and details were difficult to obtain. The retinal blood vessels were

not sharply defined and normal retinal light reflexes were absent.

The IR colour photography of the ocular fundus by means of Ektachrome-IR-aerofilm

8843 was superior to IR black-and-white film because it permitted choroidal changes,

especially the choroidal tumours, to become clearly prominent.129 Kodak Ektachrome

IR colour film has three emulsion layers sensitized respectively to peak in the green, red

and infrared with a combined sensitivity spectrum of 350 to 900 nm. A yellow filter is

used on the camera to withhold the blue to which each of these layers is also sensitive.

Upon processing the resulting colour display is a modified “false-colour” rendition of

the subject (Table 2).


- 23 -

Table 2. The resulting colour display of infrared colour photography of the ocular

fundus.35,40

Subject Colour

Healthy retina yellow-red

-retinal vessels pale shadows

-haemorrhages pale shadows

-exudates pale shadows

Arterial occlusion (post) cyan blue

Coloboma blue

Melanotic lesion of fundus

-retinal layer red-brown

-choroidal layer dark blue

-sclera reddish

The modified IR colour photography of the ocular fundus was popular during the first

half of the 1970s, but since then is rarely used due to difficulties in acquiring IR colour

film.

2.1.10 Fluorescein angiography

Fluorescein dye was synthesized by Adolf von Baeyer in 1871.10 Fluorescein staining

found early use in the diagnosis of various corneal disorders.28 Ehrlich revealed that

fluorescein passes through the blood-aqueous barrier and can be traced in the aqueous

humour after a subcutaneous injection in rabbits.65 Already in the early decades of the

20th century the first attempts to use fluorescein for in vivo staining of the blood vessels

of the iris and of the ocular fundus occurred.105,125 When Novotny and Alvis reported in


- 24 -

1960 and 1961 that retinal vessels could be photographed after an intravenous

fluorescein injection, FAG became an important method of examination in

ophthalmology.166,167

Sodium fluorescein (C20H10Na2) is a highly watersoluble organic molecule with a

molecular weight of 376.27.164 Excitation of fluorescein occurs when it is exposed to

blue wavelengths between 465 and 490 nm, resulting in the emission of yellow-green

frequencies (520 to 530 nm).186 In the bloodstream of the ocular fundus fluorescein is

excited by a wavelength of 465 nm to the maximum excitation wavelength of 500nm,

and emits a wavelength of 525 nm.46 In practice, light energy necessary for excitation is

projected into the globe by using a blue excitation filter. The emitted light energy is

documented by using a green emission filter in front of the camera.

After intravenous injection, the fluorescein is equally distributed within 3 to 5 minutes

throughout the blood. Only 60% to 80% of sodium fluorescein is bound to plasma

proteins, particularly to albumin.22 The free dye passes readily across the capillaries and

enters the tissues. Normal retinal capillaries and RPE do not leak fluorescein whereas

loss of their integrity permits leakage of free fluorescein into the neurosensory retina.

New vessels in the retina and iris leak profusely. Most of the fluorescein is eliminated

from the bloodstream within one hour, predominantly by the kidneys.186

FAG is an excellent tool for studying the retinal circulation but it has limitations with

regard to the choroidal circulation. The blue-green fluorescein excitation wavelengths

are absorbed and scattered by RPE and macular xanthophyll, producing the “black

macula” seen in FAG. The free fluorescein leaks rapidly out of the fenestrated


- 25 -

choriocapillaris, producing the diffuse background fluorescence that obscures

visualization of the deeper choroidal vessels.

2.1.10.1 Fluorescein angiography of the ocular fundus

In 1959, MacLean and Maumenee studied the human fundus by angioscopy after

intravenous injection of fluorescein.145 The development of FAG of the ocular fundus

created a revolution in the understanding and treatment of posterior segment disease.167

The classic description of “The pathogenesis of disciform detachment of the

neuroepithelium” by Gass in 1967, his stereoscopic atlas of macular diseases and its

subsequent revisions set new standards in fundus diagnosis.86,88 Several textbooks on

FAG of the ocular fundus by using film-based images for analysis of fundus disorders

gave basis for clinical research, teaching, and patient care for diabetic retinopathy,

branch vein occlusion, AMD and other retinal and choroidal

diseases.4,33,148,170,188,191,206,217,246,249 The Finnish ophthalmologists also contributed to the

FAG studies on the ocular fundus.111,113,114,137,138,173,194,195,201,235 Already in 1978 dozens

of books and thousands of articles had been written about FAG, and the technique had

become standardised and routine.103 All medical-retinal specialists and later on nearly

every ophthalmologist were able to interpret film-based FAG images as a guide for

treatment. Different applications of laser treatment have been based almost entirely on

the results of this examination technique.147

2.1.10.2 Fluorescein angiography of the iris

In 1968, Jensen and Lundbaek were the first to use fluorescein photography of the

iris.120 By using a modified fundus camera on patients with diabetes they found

fluorescein leakage from the pupillary border. This finding was corroborated by


- 26 -

Baggesen in 196911 by using a modified slit-lamp camera developed by Bruun – Jensen

for iris angiography.25 Raitta and Vannas used a slit-lamp camera for iris

angiography.174 They observed vasodilatation and neovascularization of the iris after a

central retinal vein thrombosis. Vannas modified a Zeiss Opton slit-lamp for iris

angiography.233 He observed vascular proliferation and fluorescein leakage in the iris of

patients with pseudoexfoliation and capsular glaucoma. For stereoangiographies the

equipment was equipped with the necessary angle adjustment device.

In 1974, Helve and Nieminen developed the technique for simultaneous bilateral FAG

of the anterior eye.102 By using this method Saari et al. observed an ischaemic sector,

narrow radial vessels, neovascularization, and fluorescein leakage in the iris of patients

with Fuchs’s heterochromic cyclitis.198

2.1.10.3 Anterior segment fluorescein angiography in corneal diseases

The first report concerning FAG of the limbal vasculature came from Mitsui et al., who

demonstrated the vascular patterns of normal and inflamed limbus and of progressive

and regressive pannus.156 FAG of the anterior segment was used to study the corneal

vascularization in trachoma, herpes simplex keratitis, lipid keratopathy and unsuccessful

penetrating graft, and in rosacea keratitis.63,133 Saari studied the vascular changes in

inflammatory diseases of the cornea by means of anterior segment FAG.193 He observed

that simple avascular central and marginal corneal ulcers stained with fluorescein in the

late phase of angiography, progressive pannus with pronounced fluorescein leakage

occurred in chronic corneal ulcer, disciform keratitis, Mooren’s ulcer, and complicated

acute keratoconus whereas in sclerokeratouveitis and in gutter associated with

rheumatoid arthritis the corneal vessels showed less leakage.193


- 27 -

2.1.11 Digital imaging of the eye

The CCD, an image sensor containing an array of light sensitive capacitors, was

invented in 1969. The CCD measures the amount of light on a certain surface area and

turns that information into electrical current.70 The signal in the current is put through

analog-to-digital conversion and thus the unique original analog image becomes a

perfectly reproducible digital image. The size of the active area, signal-to-noise ratio,

contrast resolution and spatial resolution are characteristics that are common to all

CCDs.70 The size of the active area determines how large the original image can be.

This can be modified by creating a smaller or larger array of capacitors. The signal-to-

noise ratio is a measurement of the relationship between meaningful information and

meaningless background information.204 Imperfections in the image capture system can

never be completely eliminated. The contrast resolution signifies the capability of the

system to measure different intensities of light. It can be modified e.g. by allowing the

capacitors more time to receive signals. Spatial resolution is the distance between points

of measurements and can be modified by increasing the number of capacitors in a given

area.232 Most optical systems can be modified to create a digital record by adding a

charge-coupled system and a computer to store the record.

The development of ophthalmic video digitizing systems began in the early 1980s. PAR

Microsystems (Atlanta, GA), produced in 1983, was the first marketable ophthalmic

imaging system (IS-2000) with the video images digitized to a 512 x 512 pixel

resolution.203 There were 24 ribbon cables with more than twelve hundred connections

for just the video capture and memory boards. By using this equipment the image

sharpening took 2-3 minutes and image registration took over 5 minutes. The cost of the

basic system was 150 000 US$ and the disks approximately 400 US$ each.203


- 28 -

In 1987, Topcon purchased the PAR Microsystems and released the product under the

IMAGEnet name.203 In 1990, they integrated the Kodak Megaplus (Model 1.4) high

resolution video camera into their use. In 1995 Topcon brought to ophthalmic

photography the Microsoft Windowsbased software, transferring the images onto

computer networks.203

A Zeiss-adapted fundus camera using a Kodak (Rocherster, NY) digital camera was

introduced in 1988. Later Zeiss models used the Kodak DCS-100 and DCS-200 camera

digital backs.203

In ophthalmology, digital images began to supplant film in FAG during the 1990s.232

Technical development of better and cheaper electronic cameras, computers, imaging

programs, and scanners have created more opportunities for the use of digital imaging

techniques in ophthalmology.211 Currently ICG angiography and OCT are digital

imaging methods that are widely available.251

2.1.12 Indocyanine green angiography

The choroidal circulation can be studied by ICG angiography which utilizes near-IR

light wavelengths and ICG dye. Near-IR light is less absorbed than visible light by the

RPE and the macular xanthophyll. Approximately 98% of ICG dye is bound to plasma

proteins and therefore it does not leak from the choriocapillaris as sodium fluorescein

dye typically does.32 ICG is considered non-toxic. It is removed from the blood stream

by the liver within the first minutes following intravenous injection.247 ICG dye absorbs


- 29 -

light in the near-IR region of the spectrum (maximum absorption at 790-805 nm). ICG

fluoresces in the near-IR region with the maximum emission occurring at 835 nm.78

The first succesfull choroidal ICG-absorption angiograms after injection of the dye into

the carotid artery of a monkey or human patients showed the choroidal veins.42,130 The

first successful choroidal ICG-absorption angiography in humans by using intravenous

injection of ICG dye and IR sensitive black and white film was reported by Flower and

Hochheimer in 1972.77 With this method the largest choroidal veins could be visualized

but the choroidal arteries and choriocapillaris could not be seen. By ICG-fluorescence

angiography both the choroidal arteries and veins could be photographed.76 On the

photographic prints the vessels appeared white and the background black.

The early clinical ICG-fluorescence angiography studies were only performed in a few

European countries.34,38,112,193 The poor sensitivity of IR film and lack of commercially

available equipment prevented widespread use of film-based ICG angiography.20

The use of improved and commercially available digital high-speed video angiography

equipment increased the interest in ICG angiography.251,252 ICG angiography was used

to study the normal choroidal circulation.75 It was especially needed for the new

treatments of AMD associated CNV, particularly occult forms, and location of feeder

vessels that carry blood to CNV membranes.82,180,251

2.1.13 Scanning laser ophthalmoscopy

In the SLO a narrow 1 mm beam of a low power laser traverses the optical axis to a

single 10 µm point of the fundus and the image is generated by scanning the retina in a


- 30 -

raster fashion. The reflected light is recorded by a photodiode and the digitised image is

stored in a PC linked to a monitor.30,241, 251 Observation at a variety of wavelengths,

including the IR, is possible by changing the laser and detector.30 By using SLO the

patient can tolerate the procedure well because it does not involve bright flashes, the IR

wavelengths of the SLO can penetrate media opacities and the imaging enables better

resolution as compared to conventional cameras.30,251

In the confocal SLO a small pinhole is moved in front of the photodiode on a conjugate

plane to the retina to select light reflected from different focal planes in the retina and

reveal three dimensional aspects of structures.240 Rodenstock (Rodenstock, Ottobrunn,

Germany) provided facilities for confocal imaging, FAG and ICG angiography.

In high-speed angiography the acquisition rate is higher than one image per second, the

maximum rate at which the Xenon flash lamp in most early fundus cameras could be

fired during FAG.251 Commercially available SLO permits ICG angiography image

acquisition at rates up to 16 images per second. The main clinical application of high-

speed ICG angiography has been in identification of CNV feeder vessels.251

2.1.14 Optical coherence tomography

In 1991, Huang et al. described a new technique, OCT, for ophthalmic structural

imaging.109 OCT performs high-resolution, micron-scale, cross-sectional, or

tomographic imaging of the internal microstructure in biological tissues by measuring

the echo time delay and intensity of backscattered or backreflected light.84,99,109 OCT

enables real-time in situ imaging of the tissue structure with a resolution of 1 to 15 μm,

which is one to two orders of magnitude finer than such imaging technologies as


- 31 -

ultrasound, magnetic resonance, or computer tomography. OCT can be performed

without physical contact to the eye, thereby minimizing patient discomfort during

examination.84

2.1.14.1 OCT retinal imaging

OCT has had its largest clinical impact in ophthalmology for retinal imaging. The

instrument design for OCT retinal imaging is similar to a fundus camera. A high-power

objective lens is used so that the retina is relay imaged onto an image plane inside the

instrument. A video camera enables real-time viewing of the retinal image. A standard

instrument has an ~30 degree field of view. OCT imaging can be performed at different

locations of the fundus by controlling the OCT scanning beam with a computer.207

OCT imaging of the retina is performed with IR light at an ~800 nm wavelength. The

image is displayed using a false colour or gray scale map that corresponds to detected

backscattered light levels ranging between 4x10-10 to 10-6 of the incident light.84

OCT enables cross-sectional images of the retina with micrometer-scale resolution. It

can discriminate the cross-sectional morphologic features of the fovea and optic disc,

the layered structure of the retina, such as the nerve fibre layer, ganglion cell layer, or

photoreceptors, and normal anatomic variations in retinal and retinal nerve fibre layer

thickness with 10 μm depth resolution.99 OCT was used to study different macular

diseases, including macular oedema, macular holes, central serous chorioretinopathy,

AMD and CNV, and epiretinal membranes. OCT has been used for monitoring of

glaucoma and macular oedema in diabetic retinopathy.207


- 32 -

2.1.14.2 OCT anterior segment imaging

In 1994, Izatt et al. first demonstrated OCT imaging of the anterior eye.117 The axial

image resolution was 10 μm and it was first performed with IR light at an ~ 800 nm

wavelength. The image spanned a transverse dimension of 21 mm, thus allowing cross-

sectional imaging of the entire anterior chamber. It showed the curvature of the anterior

and posterior surfaces of the cornea, the depth of the anterior chamber, and a cross-

section of the iris. The iris scattered light strongly, so deeper structures were

shadowed.117

The instrument design for OCT anterior eye imaging was similar to a slit-lamp

biomicroscope. The OCT beam can be positioned and scanned under operator control so

that it generates real-time cross-sectional images of selected structures in the anterior

eye. For imaging the anterior segment, a large depth of field is required, thus the

focused spot size of the OCT beam must be large. As a result, the OCT image resolution

in the transverse direction is usually coarse. Therefore imaging of the anterior eye was

later on typically performed using 1310 nm wavelengths.172 Longer wavelengths are

scattered less and they enable deeper image penetration depths into the anterior

segment. Since these wavelengths are not visible to the operator using conventional

video cameras, a visible aiming beam is used to guide the positioning of the OCT

beam.84

The rapid popularization of corneal refractive surgery spurred investigators to apply

OCT to corneal imaging.108,131 They used high-speed imaging at 1300 nm wavelength

for corneal and anterior segment OCT. They studied the cornea after LASIK operation,

differential diagnosis of keratoconus and post-LASIK keratectasia, anterior chamber


- 33 -

width and other biometric parameters, surgical results after trabeculectomy and deep

sclerectomy, and some ocular pathologies of the anterior eye.108

2.2 Teleophthalmology

Telemedicine ie. the transfer of electronic medical data from one location to the other

has existed since at least the 1960s when the National Aeronautics and Space

Administration of the United States was forced to develop methods to follow the health

of astronauts in space from the ground. In one of the earliest telemedicine projects

resulting from this pioneering work the National Aeronautics and Space Administration

collaborated with Indian Health Service of the United States government to send a

mobile health unit into the midst of the Papago tribe in Arizona USA.83 Two-way

television, radio and remote telemetry was linked to an Indian Health Service hospital

and the mobile health unit could be transported to remote areas of the reservation. The

definitive benefit of telemedicine is to bring medical services to isolated, geographically

dispersed, and physically confined persons unable to reach a physician within

reasonable time or distance.14 Technological advancements in high-resolution digital

imaging, computers, and communication networks are constantly advancing the

opportunities and improving the quality of telemedical consultations.39 Having medical

images in a digital form advances modularity which helps in the creation of

standardised tools for magnifying, analyzing, enhancing, archiving, transferring and

printing copies of the images.255

In ophthalmology diagnoses of glaucoma, diabetic retinopathy, and diseases of the

anterior segment are often based on images.254 This provides opportunities to apply

telemedicine in ophthalmology. Engagement in an efficient teleophthalmological


- 34 -

program requires that the examination results be converted into a digital form and that

ophthalmoscopy and other analog examination methods be replaced or fitted with a

system that enables the examiner to capture a digital record.30,218 Telemedicine has been

used successfully in the creation of cost-effective screening methods and programs for

ophthalmological diseases.140,169

2.3 Grading of diabetic retinopathy

In 1856, Jaeger was the first to describe diabetic macular changes in the form of

yellowish spots and extravasations of the retina.118 The meaning of these findings was

controversial until it was confirmed histopathologically that diabetes was linked to

macular oedema.163 The first case of proliferative retinopathy was reported by Manz in

1876.151 When insulin treatment for diabetes became standard, the numbers of advanced

cases of DR grew quickly. Retinal photocoagulation was discovered by Meyer-

Schwickerath in 1949 and further developed in the 1950s and 1960s as a method to

prevent diabetic neovascularization.44 Simultaneously the development of FAG in the

early 1960s made new information available on the pathological processes of DR. The

Airlie House Symposium of 1968 set the stage for important developments like

photocoagulation of wide areas of the retina, defined the relevant clinical questions of

the era and created the basis for the Airlie House classification of DR.15

A number of multicenter studies such as the Diabetic Retinopathy Study and the

ETDRS report No. 10 took up the challenge of developing the standarised classification

of retinopathy suggested by the Airlie House symposium.50,61 These important

multicenter studies verified the meaningfulness of various retinal characteristics that had

been suggested to be related to diabetes.


- 35 -

Fig. 1. Seven standard fields of the modified Airlie House classification (shown for a right eye). Field 1

is centered on the optic disc and field 2 on the macula. Field 3 is temporal to the center of macula. Fields

4 to 7 are tangential to horizontal lines passing through the upper and lower poles of the disc and to a

vertical line passing through its center.

The ETDRS report No. 10 established a grading procedure in which nonsimultaneous

stereoscopic pairs of seven standard fields (Fig. 1) of the retina were taken with a 30º

fundus cameras.61 Certain characteristics were graded only if they appeared in the

photographic field containing the optic disc: new vessels on or within 1 disc diameter of

the disc, dilated tips of new vessels of the disc, fibrous proliferation on or within 1 disc

diameter of the disc, plane of proliferation on or within 1 disc diameter of the disc and

papillary swelling. Characteristics that were only graded in the macular field were the

hard exudates and exudate rings, posterior vitreous detachment, retinal thickening

including the size of the thickened area and maximum thickness of the thickened area,

cystoid spaces, and clinically significant macular oedema. Other characteristics were

included when occurring in any of the photographic fields in some cases discluding the

fields 1 and or 2: microaneurysms; haemorrhages; drusen; hard exudates; soft exudates;

intraretinal microvascular abnormalities; venous abnormalities such as venous beading,

narrowing, loops, duplication, sheathing or perivenous exudates; arterial abnormalities

such as arteriolar narrowing or arteriolar sheathing; arteriovenous nicking; new vessels


- 36 -

elsewhere than in the retina; dilated tips of new vessels elsewhere; fibrous proliferations

elsewhere; plane of proliferation elsewhere; preretinal haemorrhage; vitreous

haemorrhage; retinal elevation or scars of prior photocoagulation. A fourth group of

nondiabetic abnormalities were also graded when noticed: questionable abnormality,

asteroid hyalosis, central vein occlusion, branch vein occlusion, central artery occlusion,

branch artery occlusion, disciform macular degeneration, chorioretinal scar not

including photocoagulation scars, choroidal nevus, subretinal fibrous tissue, coloboma

or other or questionable abnormalities.

The grades for each feature could be converted into descriptive terms of the feature:

absent, questionable, definitely present, moderate, severe and very severe. These grades

were further summarised by noting the maximum grade for a feature in any

photographic field and the number of non-overlapping fields in which the characteristic

was present.61

The results of the ETDRS study suggested that variability in the grading process was

the smallest when grading new vessels and fibrous proliferations elsewhere than in the

retina, haemorrhages, microaneurysms, hard exudates, retinal thickening and clinically

significant macular oedema with substantial interobserver agreement.61 The study went

on to divide characteristics of retinopathy into groups according to their severity and

suggest a scale of progression of DR.43

The grading of DR is important in relation to screening for DR. WHO set the criteria for

a screenable disease in 1971: the disorder that should be screened is well defined,

knowledge is available about the prevalence and the progression of the disorder,


- 37 -

effective treatment modalities exist, simple and safe screening methods are available

and the screening programme is cost effective.250 Diabetic retinopathy fulfills these

criteria and recommendations for starting screening programs were published in various

parts of Europe.122

Alternatives to the ETDRS grading protocol were devised due to the impracticality of

the original protocol in relation to its use on a large scale and the development of new

cameras with a greater than 30º photographic field. Experimenting with different

numbers and sizes of photographic fields showed that enlarging the screened area of the

retina resulted in significantly more neovascularization discovered.244 Retinal

photography has the methodological advantage of creating a document for repeated

evaluations and comparisons during follow-ups. Digital photography of the retina adds

the possibility of sending the photographs through a network to a remote location in

which experts can quickly analyse the photographs. This enables a screening

programme to reach parts of the population that were previously left unscreened.

2.4 Melanocytic tumours of the uvea

Uveal melanoma arises from neuroectodermal melanocytes within the choroid, ciliary

body or iris.6 In1905, De Schweinitz suggested that melanomas of the eye may be nevi

that have turned malignant, currently this is the opinion of most ophthalmologists.208

Uveal melanoma is the most common primary malignant neoplasm of the eye in

Caucasian adults and uveal nevus is the most common benign tumour.6,7 Uveal nevi

are usually asymptomatic whereas uveal melanomas tend to grow, invade, metastasise

to the liver and other tissues, and cause death.


- 38 -

2.4.1 Choroidal nevus and melanoma

The typical melanocytic choroidal nevus is a small gray to brown tumour usually with a

bland surface appearance. The tumour may exhibit surface alterations such as drusen

and retinal pigment epithelial clumping and typically has a basal diameter smaller than

5 mm and thickness of 1 mm or less.6,7

Diagnosis and management of small choroidal melanoma remains under heavy

discussion.221 A number of benign and malignant lesions can simulate the

ophthalmoscopic appearance of choroidal melanoma.36,216 Careful repeated observations

are recommended as a method to document growth, the primary differentiating feature

of a small choroidal melanoma from a choroidal nevus. Features that have had an

association with the risk of tumour growth include the presence of drusen, orange

pigment, subretinal fluid, retinal pigment epithelium changes, juxtapapillary location

and symptoms.8,26,37,85,155,214 Malignant melanomas may occasionally be only 2 or 3 mm

in diameter showing clinical similarity to benign choroidal nevus.155, 212 Studies on

tumour doubling times have indicated that metastasis from choroidal melanoma can

occur early in the course of the disease, when the tumour is about 3 mm in basal

diameter and 1.5 mm in thickness.68

Many of the pseudomelanomas of the ocular fundus may be differentiated from

malignant melanomas on the basis of their clinical appearance, fluorescein angiographic

features, and ultrasound characteristics, but some lesions may present diagnostic

difficulty.212,213 FAG alone is not a reliable method for differentiation between choroidal

nevus and small choroidal melanoma and between neoplasms and simulating lesions of

the ocular fundus.29,87,154 In a prospective study on 51 patients with histologically


- 39 -

confirmed choroidal melanomas FAG was diagnostic for melanoma in 63%, equivocal

in 6%, and nondiagnostic in 30% of all the cases.29

2.4.2 Nevus and melanoma of the iris

The typical melanocytic tumour of the iris appears as a localised melanotic stromal

lesion and may involve any portion of the iris. Typically the diameter is 3 mm or less

and thickness is 0.5 mm or less.6,7

A correlation exists between light iris colour and iris melanomas.128,187 In Finland one

out of every 20 uveal melanomas is a primary melanoma of the iris.175 Uveal melanoma

of the iris has less potential to metastasise and kill the host than melanoma of the

choroid or ciliary body due to early detection and treatment of melanomas of the

iris.6,124 Patients with malignant melanoma of the iris are on average ten to 20 years

younger than those with a similar tumour of the choroid or the ciliary body.225

Melanoma of the iris is equally probable to occur in either eye but bilateral tumours are

almost unknown. The neoplasm is more likely to occur in the lower half of the iris.

In the management of melanocytic lesions of the iris it is normal to start with periodic

observation to observe the possible enlargement of the mass.228 These tumours may ring

the iris, distort the pupil, undergo necrosis, cause secondary iritis, and increase

intraocular pressure when there is extensive angle involvement.89 They may involve

focal ectropion or pupillary peaking. If pronounced growth is noted, complete excision,

usually by a sector iridectomy is suggested.


- 40 -

Fluorescein leakage has been observed at the site of iris melanomas and a number of

studies have reported disorganised leaking vessels in iris melanomas.13,31,41,48 Other

studies dispute the presence of disorganised vessels as an important characteristic in the

differential diagnosis between benign and malignant iris tumours.89,132 Neither

transpupillary nor diascleral transillumination, by using white light, has been helpful in

diagnosing iris tumours, since the pigment epithelium of the iris forms a barrier to the

transmission of visible light.7

2.5 Vogt-Koyanagi-Harada disease

Vogt in 1906 and Koyanagi in 1929 reported cases with symptoms of bilateral

nontraumatic chronic iridocyclitis associated with poliosis, alopecia, vitiligo and

dysacusis.134,237 In 1926, Harada described a variation of this syndrome consisting of

bilateral posterior uveitis with an exudative retinal detachment and pleocytosis of the

cerebrospinal fluid.96 The connection between these two forms of the disease was

suggested by Babel in 1932.9 In current literature the disease is generally known as

VKH syndrome but is sometimes referred to with other names such as

uveomeningitis.168

VKH disease generally occurs in people with a dark pigmentation such as orientals,

american indians, hispanics and blacks.81 Currently it is thought to be caused by a

T-lymphocyte mediated autoimmune reaction targeting melanocytes.90 Shindo et al.

suggested an association with HLA alleles DRB1*05 and DRB1*10.219

The American Uveitis Society established the criteria for VKH syndrome in 1978 as

follows: The presence of at least three of the following findings: bilateral iridocyclitis,


- 41 -

posterior uveitis with exudative retinal detachment or sunset glow fundus, central

nervous system involvement and dermatologic manifestations such as alopecia, vitiligo

or poliosis in the absence of previous ocular trauma or surgery.80 The criteria were

revised at the First International Workshop on VKH disease in 1999, with allowance for

the different ocular findings present in the early and late stages of the disease.177

The progress of VKH disease is generally divided into four stages, a prodromal stage

consisting mainly of auditory or neurologic symptoms, an acute uveitic stage, a chronic

phase characterised by the variable development of depigmentation, and a

chronic-recurrent phase with iridocyclitis.177 An early administration of high-dose

corticosteroid treatment generally turns VKH disease quiescent but does not stop the

development of chronic-recurrent uveitis in all cases.116,189


- 42 -

Aims of the Present Study

- 43 -

3. AIMS OF THE PRESENT STUDY

The purpose of our study was to:

1. Test the sensitivity and specificity of digital retinal images in grading DR.

2. Evaluate comparative use of digital colour, red-free, and red light imaging in the

differential diagnosis of choroidal nevus and melanoma.

3. Develop and assess digital IRT imaging of the iris.

4. Describe the changes of the iris and ciliary body in the chronic phase of VKH

disease by using digital IRT imaging.

5. Evaluate IRT imaging and FAG findings in differential diagnosis of iris nevus

and melanoma.

Aims of the Present Study

- 44 -

Materials and Methods

- 45 -

4. MATERIALS AND METHODS

4.1 Subjects (I – V)

Altogether the material comprised 398 subjects including 154 men and 244 women with

ages ranging between 12 and 93 years. The total number of eyes examined was 469

(Table 3).

In the study on the sensitivity and specificity of digital retinal images in grading DR (I)

we included 70 subjects, 37 men and 33 women, with an age range between 22 and 83

(mean 41.8 ± 19.1) years. Forty-two of the subjects, 28 men and 14 women were

patients with diabetes (79 examined eyes) aged between 23 and 83 (mean 53.2 ± 16.8

years). The mean duration of the disease was 19.8 ± 12.2 (range 1 – 44) years. Forty

patients had DR and 31 of them had been treated with photocoagulation (56 examined

eyes). The control group consisted of 28 healthy medical students (29 examined eyes)

with an age range between 22 and 31 (mean 24.7 ± 1.7) years.

For the study on digital imaging in differential diagnosis of small choroidal melanoma

(II) we re-evaluated the case records of 241 consecutive patients referred to the Turku

University Eye Clinic for suspected choroidal melanoma. Eighty three of the patients

were men and 158 women, with the ages ranging between 13 and 93 (mean 64 ± 14)

years. The entry criteria for the digital imaging study were digital colour, red-free and

red light imaging of the ocular fundus due to a suspected small choroidal melanoma. Of

the 241 patients, 110 (45.6%), 33 men and 77 women, with an age range between 15

and 91 (mean 66 ± 15) years, fulfilled the inclusion criteria for the digital imaging

study.


- 46 -

Table 3. Number of examined subjects and eyes in different diagnostic groups

Original Diagnostic Number of Number of

publication group examined subjects examined eyes

I Diabetic 42 79 patients

Controls 28 29

Total 70 108

II Suspected 241 241 choroidal melanoma

III Normal 29 59 subjects and eyes

Blunt eye trauma 1 1

Total 30 60

IV VKH disease 1 2

V Melanocytic 56 58 tumours of iris

Total 398 469

To describe a new method for digital IRT imaging of the iris (III) we studied 30

subjects (60 examined eyes), 10 men and 20 women, with ages ranging between 22 and

30 (mean 24.3 ± 1.9) years. Twenty nine of them were healthy subjects with no known

general or eye diseases. One 30-year-old male student that was otherwise healthy had

suffered a blunt sport eye injury of the right eye at the age of 12. In biomicroscopic

examination all eyes were calm with no aqueous flare or cells, the irides showed normal

structure, and the ocular fundus was normal in all volunteers.


- 47 -

To describe iris atrophy, serous detachment of the ciliary body, and ocular hypotony in

the chronic phase of VKH disease (IV) we did an ocular examination and follow-up

including digital IRT imaging of the iris and ciliary body in both eyes of a 52-year-old

woman with chronic phase of this disease.

To compare IRT imaging and FAG findings between iris nevus and melanoma (V) we

studied 56 patients, 24 men and 32 women, with an age range between 12 to 92 (mean

57) years. All of these patients had melanocytic tumours of the iris. Two patients had an

iris nevus in both eyes. Thus the material consisted of altogether 58 eyes with a

melanocytic tumour of the iris.

4.2 Diagnostic criteria (I – V)

4.2.1 Diabetes (I)

In the study on evaluation of digital fundus cameras for DR screening (I) there were 42

subjects with diabetes. Seventeen of them (40.5%) had type 1 diabetes and 25 (59.5%)

type 2 diabetes as defined by the American Diabetes Association.69

4.2.2 Grading of diabetic retinopathy (I)

The grading of DR used in this study was a modified version of the Diabetic

Retinopathy Study Report No. 8 and ETDRS Reports Nos. 7, 10 and 18 and comprised

nine diagnoses:43,50,60,61

(1) No DR. May include retinal haemorrhages, cotton wool spots and changes related to

branch vein occlusion, age-related macular degeneration, or other retinal diseases but no

MAs.


- 48 -

(2) Minimal NPDR. MAs only (usually <3-5), no other changes typical of DR.

(3) Mild NPDR. MAs, mild retinal haemorrhages, hard exudates, cotton wool spots,

venous loops.

(4) Moderate NPDR. MAs, moderate retinal haemorrhages in four quadrants or severe

in one quadrant, mild IRMA in one to three quadrants or if mild IRMA in four

quadrants or if definite venous beading in one quadrant, then no other severe NPDR

changes.

(5) Severe NPDR (preproliferative). The 4-2-1 rule characterised by one of the

following: MAs and severe retinal haemorrhages in four quadrants (4/4), venous

beading in at least two quadrants (2/4), moderate to severe IRMA in at least one

quadrant; or the 2-1-4 rule characterised by at least two of the following: MAs and

severe retinal haemorrhages in at least two up to three quadrants (2-3/4), venous

beading in one quadrant (1/4), mild IRMA in all four quadrants (4/4).

(6) Proliferative DR. Definite new vessels in the retina or on the optic disk without

high-risk characteristics.

(7) High risk PDR. PDR with preretinal haemorrhage, fibrous tissue and other high risk

characteristics50

(8) Advanced diabetic eye disease. Advanced PDR including vitreous haemorrhages,

fibrous tissue, retinal detachment, rubeosis iridis.

(9) Ungradeable. Cannot grade due to inadequate digital imaging quality or obscuration

from vitreous haemorrhage or other abnormality.


- 49 -

4.2.3 Uveal melanoma (II, V)

4.2.3.1 Choroidal melanoma (II)

All cases of suspected choroidal melanoma underwent a careful ophthalmological

examination, including ophthalmoscopy through a dilated pupil, photography or digital

imaging of the fundus, FAG of the fundus, and B scan ultrasonography of the suspected

lesion. The suspected choroidal melanoma appeared as a dark brown to a golden solid

tumour with prominent clumps of orange lipofuscin pigment on their surface and some

with visual symptoms and secondary nonrhegmatogenous retinal detachment.6,7

The diagnostic criteria to differentiate small choroidal melanoma from choroidal nevus

were similar to those reported by Shields et al. in 2004 by using the mnemonic

“TFSOM” (to find small ocular melanoma), where T = thickness greater than 2 mm, F =

subretinal fluid, S = symptoms, O = orange pigment and M = margin touching optic

disc.211 Choroidal melanosytic tumours that display none of these factors have a 3 %

risk of turning into melanoma at 5 years and most likely represent choroidal nevi.

Tumours that display one factor have a 38% risk of growth, and those with two or more

factors show growth in over 50% of cases.212 Presumed cases of choroidal melanoma

were referred to the Department of Ophthalmic Oncology at Helsinki University Eye

Clinic, which is the only ophthalmic oncology unit in Finland. An ophthalmic

oncologist confirmed the diagnosis of most cases of choroidal melanoma in our study.

The material of melanoma patients in the digital imaging study was limited to cases

with small choroidal melanoma. We used the COMS criteria for small-sized melanoma

with a 5-16 mm largest basal diameter and 1.0 – 2.5 mm height.220 All cases of medium-

sized and large-sized melanomas were excluded from the digital imaging study.


- 50 -

Similarly, patients who had undergone treatment of the choroidal melanoma before

digital imaging of the lesion were excluded from the trial. Of the 241 patients, 110

patients (45.6%), 33 men and 77 women with ages ranging between 15 and 91 (mean 66

± 15) years, fulfilled the inclusion criteria for the digital imaging study (II).

4.2.3.2 Iris melanoma (V)

Only primary iris melanocytic tumours with more than 2/3 of the tumour occurring in

the iris were included in this study. All adenomas of the iris epithelium and melanomas

that arose in the ciliary body with secondary iris extension were excluded. Each patient

underwent a full ophthalmological examination. The follow-up period from the first to

the last ophthalmological examination varied between 1 month and 34 years (mean 6

years 11 months). The tumour was classified as a melanoma of the iris in five cases of

which four were histologically proved. The fifth melanoma patient, a 51-year old man,

had an elevated iris tumour that showed involvement of the chamber angle and growth

of the largest basal diameter from 7 mm to 10.6 mm during a 9-year follow-up, and the

tumour was clinically diagnosed as a melanoma of the iris by two ophthalmic

oncologists.

4.2.4 Uveal nevus (II,V)

Uveal nevi are benign lesions composed of atypical melanocytic cells that are benign by

histopathological criteria.7

4.2.4.1 Choroidal nevus (II)

Key features of benign choroidal nevus were gray to brown (melanotic) or tan

(amelanotic) choroidal mass usually 5 mm or less in diameter and 1 mm or less in


- 51 -

thickness. Most uveal nevi were asymptomatic. They showed a bland surface

appearance with some drusen and retinal pigment epithelial alterations.7 They usually

did not show any enlargement when followed by fundus photography or digital imaging

by ultrasonography.

4.2.4.2 Iris nevus (V)

The iris nevus was a brown or fleshy iris mass usually 3 mm or less in diameter.7 In

this study in 53 eyes, the melanocytic tumour was diagnosed as an iris nevus that

appeared as a localised stromal lesion of the iris involving any portion of the iris from

the pupillary margin to the iris root. The diagnosis was based on histopathological

examination after sector iridectomy in four eyes, and in 49 eyes it was based on clinical

serial re-evaluation and the fact that the lesion did not show any growth when observed

carefully through the use of slit-lamp colour photographs during a mean follow-up of 7

years (V).

4.2.5 Vogt-Koyanagi-Harada disease (IV)

The diagnostic criteria of VKH disease used in this study were the same as reported by

the International Committee on VKH Disease Nomenclature.177 In complete VKH

disease criteria 1 to 5 must be present:

1. No history of penetrating ocular trauma or surgery preceding the initial onset of

uveitis.

2. No clinical or laboratory evidence suggestive of other ocular disease entities.

3. Bilateral ocular involvement (a or b must be met, depending on the stage of disease

when the patient is examinded)

a. Early manifestations of the disease.


- 52 -

(1) Diffuse choroiditis (with or without anterior uveitis, vitreous inflammatory reaction,

or optic disk hyperemia), which may manifest as one of the following:

(a) Focal areas of subretinal fluid, or

(b) Bullous serous retinal detachments.

(2) With equivocal fundus findings; both of the following must be present:

(a) Focal areas of delay in choroidal perfusion, multifocal areas of pinpoint leakage,

large placoid areas of hyperfluorescence, pooling within subretinal fluid, and optic

nerve head staining by FAG, and

(b) Diffuse choroidal thickening, without evidence of posterior scleritis by

ultrasonography.

b. Late manifestations of disease.

(1) History suggestive of prior presence of findings from 3a, and either both (2) and (3)

below, or multiple signs from (3):

(2) Ocular depigmentation (either of the following manifestations is sufficient):

(a) Sunset glow fundus, or

(b) Sugiura sign (perilimbal vitiligo)

(3) Other ocular signs:

(a) Nummular chorioretinal depigmented scars, or

(b) Retinal pigment epithelium clumping and/or migration, or

(c) Recurrent or chronic anterior uveitis

4. Neurological / auditory findings (may have resolved by time of examination).

a. Meningismus (malaise, fever, headache, nausea, abdominal pain, stiffness of the

neck and back, or a combination of these factors; headache alone is not sufficient to

meet definition of meningismus, however), or

b. Tinnitus, or


- 53 -

c. Cerebrospinal fluid pleocytosis.

5. Integumentary finding (not preceding onset of central nervous system or ocular

disease).

a. Alopecia, or

b. Poliosis, or

c. Vitiligo.

4.3. Classification of iris colour (III, V)

The iris colour was assessed clinically and from colour photographs of both eyes. In the

study on digital IRT imaging of the iris (III) we used the five-grade classification of

Seddon et al. by using standard photographs for grading iris colour.209 In our study (III)

73.3% of eyes belonged to iris colour grade 1 of Seddon et al.209and only eight subjects

(16 eyes) showed more brown irides (Table 4).

Table 4. Classification of iris colour of 60 eyes according to standard photographs of

Seddon et al.208 (III)

Gender Classification of iris colour Total number of eyes

1 2 3 4 5

Males 16 2 2 - - 20

Females 28 2 - 6 4 40

Total 44 4 2 6 4 60

In the study on IRT imaging and FAG of iris nevus and melanoma (V) the iris colour

was graded as light when the iris was blue, gray, or green and as dark when it was hazel,

brown or dark brown. In that study (V) the iris colour was light in 56/58 eyes (96.6%)

and hazel in two eyes with an iris nevus (3.4%).


- 54 -

4.4 Imaging methods (I-V)

4.4.1 Digital retinal imaging (I, II, IV)

All images were captured by a professional photographer at the Turku University Eye

Clinic. For all types of digital retinal imaging, the pupils were dilated with 0.5%

tropicamide and with 10% phenylephrine eyedrops.

4.4.1.1 Digital retinal imaging in screening for DR (I)

A Topcon TRC 50 IA fundus camera (Topcon Corporation, Tokyo, Japan) was used to

take two 50 º digital colour images (768x576 pixels) and one red-free, black and white

image (1320x1032 pixels) using the green filter supplied by the manufacturer. A Canon

CR6-45NM digital fundus camera (Canon Europe, Amstelveen, the Netherlands) was

used to take two 45 º digital colour images per eye (2160 x 1440 pixels). In the 45-50 º

images one field covered the temporal area, including the macula and disc, and the

second field covered the nasal area, including the disc. A commercial version of a new

handheld digital video camera for eye examination (MediTell, Medimaker Ltd, Ranua,

Finland) was used for digital colour imaging of the central parts of the ocular fundus

(768 x 576 pixels). Two single (still) images were taken using a white light flash for

illumination and stored on a PC. Captured JPEG images had a field area of about 20 º

and a resolution of 72 dpi.192

Digital retinal imaging was carried out on 108 eyes (generating a total of 427 images).

Of the 108 eyes, 106 underwent digital 50 º retinal colour imaging and 106 eyes

underwent red-free imaging (Topcon TRC 50 IA, 290 images); 104 eyes were imagined

under both of these imaging modalities. Digital 45 º retinal colour imaging (Canon


- 55 -

CR6-45NM) was carried out on 29 eyes (54 images). The handheld digital colour video

camera (MediTell) was used for the imaging of 44 eyes (83 images).

4.4.1.2 Digital retinal imaging of choroidal melanocytic tumors (II)

A Topcon TRC 50 IA fundus camera (Topcon Corporation, Tokyo, Japan) was used to

take digital colour images (1360x1024 pixels), red-free black and white retinal images

(1320 x 1032 pixels), and red light black and white choroidal images (1320 x 1032

pixels) using the filters supplied by the manufacturer.

4.4.1.3 Digital retinal imaging of the Vogt-Koyanagi-Harada patient (IV)

A Topcon TRC 50 IA camera (Topcon Optical Co. Ltd., Japan) was used for digital 50º

colour (768 x 576 pixels), blue light, red-free, red light, and fluorescein angiography

imaging of the ocular fundus. The red-free light, monochrome, black and white images

(1320 x 1032 pixels) were taken using the green filter supplied by the manufacturer.

4.4.2 Colour photography of the anterior eye (III, IV, V)

Colour photographs of the iris were taken with a standard Zeiss stereo slit-lamp camera

(Carl Zeiss, Oberkochen, Germany) (V). Digital colour images of the anterior eye were

taken using the Zeiss photoslitlamp with a digital camera body (Kodak DCS-315) (III,

IV).

4.4.3 Digital transpupillary transillumination imaging of the iris (IV)

Digital transpupillary transillumination imaging of the iris using white light was done

using the Zeiss photo slitlamp digital camera body (Kodak DCS-315) (IV).


- 56 -

4.4.4 Infrared transillumination imaging of the iris and ciliary body (III, IV, V)

For digital infrared transillumination imaging oxybuprocaine eye drops (Oftan Obucain,

Santen, Tampere, Finland) were instilled onto the eye for topical anaesthesia.

4.4.4.1 Infrared transillumination stereophotography of the iris (V)

A high-intensity light from an electronic flash unit was led through a Kodak Wratten

gelatine filter No. 87 (Eastman Kodak Company, Rochester, New York, USA) filtering

unwanted visible light and transmitting only the IR light (Fig. 2 and Table 1) via a

fibreglass optic through the lateral wall of the bulbus (Fig. 3 and 4). Stereoscopic pairs

of photographs were taken with a Zeiss cornea microscope on Kodak high speed IR

black and white film, which was sensitive to reflected radiation in the range of 700 to

900 nm.62

4.4.4.2 Digital infrared transillumination imaging of the iris and ciliary body (III,

IV, V)

The examination light from the incandescent lamp (Guerra 3746/Z 6V 30W) of the

Zeiss slitlamp was passed through a Kodak Wratten gelatine filter No. 87 transmitting

only IR light via a 190 cm long light fiber optic cable (Luxtec, West Boylston,

Massachusetts, USA) through the temporal eye wall. A Topcon retinal camera (TRC 50

IA, Topcon Corporation, Tokyo, Japan), with its indocyanine green angiography filters

in place, and focused on the iris, was used to capture the IRT images of the iris.


- 57 -

Fig. 2. Transmittance curve for Kodak Wratten filter No. 87 used for infrared photography.62

4.4.5 Fluorescein angiography (II, IV, V)

For each FAG we used 0.043 ml/kg of 10% sodium fluorescein (Fluorescite 10%, Alcon

Inc, Forth Worth).

4.4.5.1 Fluorescein angiography of the fundus (II, IV)

For fundus FAG the pupils were dilated with 0.5% tropicamide and with 10%

phenylephrine eyedrops. A Topcon TRC 50 IA fundus camera (Topcon Corporation,

Tokyo, Japan) was used for digital FAG imaging of the fundus using the filters supplied

by the manufacturer.


- 58 -

Fig. 3. IRT photography of the iris showing the illuminating probe on the lateral side of the bulbus

Fig. 4. IRT photography of the ciliary body and iris. Illuminating probe located posteriorly near equator


- 59 -

4.4.5.2 Fluorescein angiography of the iris (V)

Fluorescein angiograms of the iris were taken of 44 eyes. In 24 eyes, the analog FAG

was done using the Zeiss biomicroscope in which a motor-driven Nikon camera (Nikon

Corporation, Kawasaki, Japan) was connected. We used a Baird Atomic interference

filter B4 (Baird Atomic, Cambridge, Massachusetts, USA) and as the barrier a Kodak

Wratten filter No. 15. In 20 eyes, a digital iris FAG was taken using the Topcon TRC 50

IA fundus camera connected to a Kodak Mega Plus digital black and white camera 1.4i

focused on the iris.

4.5 Subtraction method to evaluate growth of choroidal melanocytic tumours (II)

A pair of digital colour images from different dates was chosen and compared by using

a special built-in utility tool of the Topcon Image Net Program (Topcon Corporation,

Tokyo, Japan). This tool can be used to compare the location of the borders of the lesion

using the subtraction method. At least three points of correspondence were selected in

each image to find the relative scale. Each point was first marked on the left-side image

(taken earlier) and then on the right-side image (taken later after the the follow-up). For

maximum accuracy, the corresponding points should be as distant as possible to each

other. The border of the tumour area was drawn in the left-side image by using the

mouse whereupon the program draws the line to the corresponding anatomical location

of the fundus in the right-side image (line A). The possible growth is evident if the

lesion of the right-side image extends over the line A. The amount of growth of the

tumour is better visualised when the border of the tumour area of the right-side image is

also drawn by using the mouse, whereupon the program also draws the line to the


- 60 -

corresponding anatomical location in the left-side image (line B). The area of growth is

between lines A and B.

4.6 Data collection, enrolment and experimental protocol (I, II, V)

4.6.1 Experimental protocol and grading of diabetic retiniopathy (I)

Each eye-camera combination was assigned a random code. The digital images and the

corresponding codes were sent in electronic form for assessment to three graders (A, B

and C). Thus there were altogether 1281 grading events. Two of the graders (B and C)

were ophthalmologists and one (A) was a bachelor of medicine with special training in

assessing images of DR. They were masked to all clinical and personal data of the

patients and to the grading of the other screeners.

All graders received written definitions of the stages of DR before the study began and

were asked to classify pathological findings according to this protocol. To begin with,

retinopathy of each eye was graded in a randomised and masked manner using digital

colour images and digital red-free images separately. Later we used the combined

results of the ophthalmological examination including the three-mirror lens funduscopy

and the digital colour and red-free black-and-white images of the ocular fundus as a

mixed gold standard reference against which the digital retinal colour and red-free

images were compared.

4.6.2 Eligibility criteria, enrolment and experimental protocol of small choroidal

melanoma study (II)

This study included all patients who had been referred to the Department of

Ophthalmology, Turku University Central Hospital from January 1, 1987 through


- 61 -

December 31, 2003 due to a suspected choroidal melanoma. All cases with

conjunctival, iris and predominantly ciliary body tumours were excluded. The cases

were abstracted from the computerised patient file system of the Turku University

Central Hospital by using the international classification codes for the diagnosis: ICD-9

for the years 1987-1995 and ICD-10 after 1996. Since nearly all cases of suspected

choroidal melanoma in the Hospital District of Southwest Finland were referred to the

Department of Ophthalmology, Turku University Central Hospital, evaluation of the

hospital records alone gave population based figures from the incidence of choroidal

melanoma. The population in the Hospital District of Southwest Finland decreased from

469,765 in 1987 to 433,178 in 1991, and then increased to 452,444 in 2003; during the

study period the mean population was 450,468.

By using the social security numbers of the patients with suspected choroidal

melanoma, we collected all those cases that had undergone digital color, red-free and

red light imaging of the ocular fundus. A random code was assigned to each case. Based

on the digital images we determined the diagnosis for every case in a random manner

and masked for all clinical data of the patients. Later, we used the combined results of

the ophtahlmological examination and follow-up including the mydriatic three-mirror

lens funduscopy, the assessments of the digital colour, red-free and red light images,

FAG and B scan ultrasonograpgy images, and when available, the diagnosis of the

ophthalmic oncologist as a mixed gold reference standard against which the

comparative use of the digital retinal colour, red-free and red light images were

compared to determine the sensitivity and specificity of this diagnostic method for the

diagnosis of small choroidal melanoma and different pseudomelanomas.


- 62 -

Visualisation of the lesion was graded in the digital images as follows: 1) the lesion was

not visualised because the examination light did not reach the lesion, or it transmitted it

without reflecting its contours; 2) the lesion was visualised; 3) the lesion was visualised

as darkly pigmented or black. We determined the distribution of different diagnostic

groups according to different combinations of grades of visualisation of the lesion when

digital colour, red-free and red light images were used.

4.6.3 Evaluation of the usefulness of the subtraction method (II)

To evaluate the usefulness of the subtraction method for demonstration of growth of the

melanocytic choroidal lesion (II), digital colour images of patients with coroidal nevus

and small choroidal melanoma were studied. The usefulness of the subtraction method

was graded as follows: 1) the subtraction method could not be used; 2) the method was

useful but did not show growth of the lesion; 3) the method was useful, showing growth

of the lesion.

4.6.4 Data analysis of infrared transillumination and fluorescein angiography

findings of iris melanocytic tumours (V)

According to the IRT, the tumour was graded as translucent, partially translucent, or

non-translucent to IR light. Fluorescein angiograms were divided into those in which

the tumour caused complete, incomplete or no masking of the iris vessels. The degree of

leakage from the tumour vessels was divided into four categories: none, early moderate,

late moderate, and late gross. The pattern of tumour vasculature was graded as normal

(geometric with vessels of small calibre) or disorganised (vessels of varying calibre)

(V).


- 63 -

4.7 Statistical methods (I, II, III, V)

The statistical program used in this study was SAS® 8.2 system for Windows. All

p-values lower than 0.05 were considered statistically significant.

To assess the sensitivity and specificity of digital retinal images in grading DR (II) the

sensitivity and specificity of the findings for different digital imaging modalities were

calculated per eye. The results were given with 95% confidence intervals (95% CI). The

percentage of overcalls, undercalls, and exact agreement were examined. The chi-square

test was used to determine whether the sensitivity and specificity of different imaging

modalities differed between patients treated with photocoagulation and those not treated

with photocoagulation, and to determine whether the overcalls and undercalls differed

between colour and red-free imaging. Levels of agreement between different imaging

modalities were determined by weighed κ statistic with 95% CIs. Interpretation of κ has

been defined earlier.74

In the study on digital imaging in the differential diagnosis of small choroidal

melanoma (III) the mean annual incidence of choroidal melanoma per 100,000

population was calculated by using the corrected population statistics of the Hospital

District of Southwest Finland for the years 1987-2003.222 Sensitivity and specificity of

the findings for combined use of colour, red-free and red light images were calculated

per eye. The results were given with 95% exact confidence intervals (95% CI). The

percentage of overcalls, undercalls, and exact agreement were examined. Levels of

agreement between comparative use of colour, red-free and red light images and the

reference standard were determined by using a κ statistic with 95% CI. Interpretation of

κ-values has been defined earlier.74


- 64 -

In the study on digital IRT imaging of the normal iris (IV), the effect of gender and of

the colour of the iris on the IRT findings was analysed by using the Fischer’s exact test.

To compare IRT imaging and FAG findings between iris nevus and melanoma (V)

the association between diagnosis and two ordinary scaled variables (infrared

transillumination and degree of masking in iris angiography) were analysed by an

extended Mantel-Haenszel mean score statistic.223 The relationships between the

outcome variable and other categorical variables were assessed using Fisher’s exact test.

All tests were two-tailed.

Results

- 65 -

5. RESULTS

5.1 Sensitivity and specificity of digital retinal images in grading diabetic

retinopathy (I)

Digital 50° red-free imaging (Topcon TRC 50 IA) had a sensitivity of 97.7% (95%

CI 95.8 – 99.7), two-field 50° colour imaging (Topcon TRC 50 IA) a sensitivity

94.0% (95% CI 90.8 – 97.2), and two-field 45° colour imaging (Canon CR6 – 45

NM) a sensitivity of 88.9% (95% CI 82.0 – 95.8) for detection of at least mild NPDR

when compared with the reference standard. No statistically significant differences

appeared in the specificity between digital red-free (98.9%; 95%CI 96.7-100) and

colour imaging (99.0; 95%CI 96.9-100 for the Topcon TRC 50 IA and 100% for the

Canon CR6-45NM) for detection of at least mild NPDR when compared with the

reference standard.

Digital red-free imaging showed 98.1% and digital colour imaging 95.5% (Topcon

TRC 50 IA) and 89.3% (Canon CR6 – 45 NM) exact agreement for detection of at

least mild NPDR when compared with the reference standard. Digital red-free

imaging only showed 0.3% and digital colour imaging 2.6% (Topcon TRC 50 IA)

and 9.5% (Canon CR 6-45 NM) undercalls for detection of at least mild NPDR when

compared with the reference standard. In all imaging modalities and in all graders the

overcalls varied between 0% and 1.0%. Only 1.3% of digital red-free images and

1.2-1.6% of digital colour images (Topcon and Canon) were ungradeable. The

handheld digital colour video camera (MediTell) only showed 6.9%; 95%CI 2.3-

11.5 sensitivity for detection of at least mild NPDR when compared with the

Results

- 66 -

reference standard and ungradeable images represented 92.3%. A good intergrader

agreement existed between all three graders for all four imaging modalities.

Digital red-free imaging showed the best sensitivity, digital colour imaging with

Topcon TRC 50 IA the second best and with Canon CR 6-45 NM the third best

sensitivity for detection of different diagnostic groups of DR. Although the

differences in the sensitivity between these imaging modalities were not statistically

significant. In all these imaging modalities the sensitivity decreased with an increase

of the severity of DR. The sensitivity of the handheld digital colour video camera

was statistically significantly lower than those of the other digital imaging modalities

in all diagnostic groups. The specificity of digital colour and red-free imaging

modalities (Topcon and Canon) was excellent in all diagnostic groups of DR. The

specificity of the handheld digital video camera was low and showed a wide

confidence interval because the percentage of ungradeable images was over 90% in

all diagnostic groups of DR.

When comparing the Topcon TRC 50 IA digital colour and red-free imaging

modalities the three graders did altogether 318 gradings per eye for both imaging

modalities. Both imaging modalities showed a very good agreement between the

imaging modality and the reference standard. Digital colour imaging showed 41

undercalls and 30 overcalls, and digital red-free imaging 24 undercalls and 65

overcalls. These differences were statistically significant (p = 0.000082). The

differences were most pronounced when grading mild and moderate NPDR.

Microvascular changes including MAs and IRMA were more clearly visualised in

red-free images than in colour images. Direct comparison between digital colour and

Results

- 67 -

red-free imaging, however, showed a very good agreement in detecting and grading

DR (weighted kappa 0.84; 95%CI 0.80-0.88; grade “ungradeable” deleted; table not

shown).

Of the patients with DR, 31 had been treated with photocoagulation which made the

grading more difficult. No statistically significant differences, however, existed in

the sensitivity figures of different imaging modalities between the 56 eyes, which

had been treated with photocoagulation and 23 eyes without any photocoagulation

(p = 0.28).

5.2 Digital imaging in differential diagnosis of small choroidal melanoma (II)

From January 1, 1987 through December 31, 2003 altogether 241 patients were referred for

clinical evaluation because of an opthalmoscopically visible lesion suspected to be a

malignant melanoma of the choroid (Table 3). Of these, 61 (25.3%) were found to have

choroidal melanoma. The mean annual incidence of choroidal melanoma in Southwest

Finland was 0.80 per 100 000 population. Of the total number, 180 (74.7%) were diagnosed

as having various simulating lesions (Table 5). The most frequent pseudomelanomas were

choroidal melanotic and amelanotic nevi (64.7%), disciform lesions (5.4%), congenital

hypertrophy of retinal pigment epithelium (1.7%) cases, and circumscribed choroidal

haemangioma (1.3%) (Table 5).

Results

- 68 -

Table 5. Final diagnosis of 241 patients referred because of suspected choroidal melanoma.

RPE denotes retinal pigment epithelium

Diagnosis ___________Patients_________ Percent of total Number Age (mean ± SD)

Choroidal nevus 156 65 ± 14 64.7

Choroidal melanoma 61 60 ± 13 25.3

Disciform lesions 13 71 ± 11 5.4

Congenital hypertrophy of RPE 4 42 ± 24 1.7

Circumscribed choroidal 3 53 ± 8 1.3 haemangioma

Choroidal haemorrhage 2 80 ± 4 0.8

Retinochoroidal scar following 1 66 0.4 retinochoroiditis

Macroaneurysm and retinal 1 78 0.4 haemorrhage

Total 241 64 ± 14 100.0

Table 6. Patients who were eligible for the digital imaging study. RPE denotes retinal

pigment epithelium

Diagnosis ___________Patients_________ Percent of total Number Age (mean ± SD)

Melanotic choroidal nevus 72 66 ± 14 65.5

Amelanotic choroidal nevus 15 66 ± 15 13.6

Small choroidal melanoma 7 61 ± 22 6.4

Disciform lesions 11 76 ± 7 10.0

Congenital hypertrophy of RPE 3 32 ± 9 2.7

Choroidal haemorrhage 1 83 0.9

Macroaneurysm and retinal 1 78 0.9 haemorrhage

Total 110 66 ± 15 100.0

Results

- 69 -

Of the 110 patients who were eligible for the digital imaging study seven (6.4%) had small

choroidal melanoma (Table 6). The most frequent pseudomelanomas were melanotic

choroidal nevus (65.5%), amelanotic choroidal nevus (13,6%), disciform lesions (10%), and

congenital hypertrophy of the retinal pigment epithelium (2.7%) (Table 6).

Melanotic choroidal nevi were pigmented, flat, small (generally smaller than 3 to 7.5 mm at

the largest basal diameter) and mostly circular or oval lesions. Digital colour images showed

melanotic choroidal nevi as slate grey to moderately pigmented lesions in 63 (87.5%) of the

72 cases, and as heavily pigmented in nine cases (12.5%). Red-free images did not show the

melanotic choroidal nevus in 61 (85%) cases and only revealed it in 11 cases due to the rich

or intense pigmentation of the lesion (15%). Red light images revealed the melanotic

choroidal nevus as dark or black in most cases (97%), and as dark grey in the two smallest

(0.5 mm in largest basal diameter) melanotic choroidal nevi. The most popular grading

combination (82%) of a melanotic choroidal nevus in digital colour/red-free/red light images

was: the lesion was seen (2)/ lesion was not seen (1)/ lesion was seen as darkly pigmented as

black (3).

Amelanotic choroidal nevi were hypopigmented or completely nonpigmented solid lesions

replacing the normal choroid clinically. The lesions were typically circular or ovoid in basal

configuration. They appeared in the postequatorial choroid but in no case beneath the foveola.

The amelanotic areas showed early and late hyperfluorescence in fluorescein angiography.

Digital colour images revealed them as a yellowish light but some of the lesions had very

mild melanotic pigmentation along the margins or extending towards the center of the lesion.

Digital red-free images showed the lesion mostly as light areas but in three cases only due to

rich drusen-like changes overlying the lesion. Red light images revealed the lesion as light

Results

- 70 -

areas with only occasional small dark patches. In all amelanotic choroidal nevi the

combination of grades of visualization in digital colour/red-free/red light images was 2/2/2.

Seven patients had a small choroidal melanoma as a circumscribed, localised, and elevated

choroidal lesion. Six of these patients had visual symptoms due to melanotic choroidal

melanomas, which appeared in digital colour images as brown or grey lesions with clumps of

orange pigment on their surface in four patients, subretinal fluid in three, and marginal touch

of the optic disc in two patients. The tumours ranged from 5 to 14 mm at the largest basal

diameter (mean, 7.8 mm) and 1.0 to 2.4 mm in thickness (mean, 1.7 mm). Fluorescein

angiography in the tumour area of three patients showed pinpoint leaks that increased in size.

Red-free images revealed only three of the small melanotic choroidal melanomas. In all six

cases with the melanotic choroidal melanoma the lesion appeared in red light images as

darkly pigmented or black. In one of the seven patients with small choroidal melanoma,

colour images revealed amelanotic choroidal melanoma as a yellowish-white oval lesion with

some pigment clumps and 4.5 x 5.5 mm at basal diameter. The tumour was 2 mm in thickness

in B-scan ultrasonography examination. Both red-free and red light images revealed the lesion

as a light area with some black pigment clumps.

A disciform lesion was visualised in digital colour images as a yellowish-white choroidal

lesion with dark greenish ring of hyperpigmentation, or as a focal choroidal infiltrate. Red-

free images did not reveal the greenish pigment-ring lesion clearly, but showed the focal

choroidal infiltrate as a light grey or whitish area surrounded by a shallow disciform retinal

detachment. Disciform lesions were visualized in all 11 cases in digital colour and red-free

images. In six cases the disciform lesion was not clearly seen in red light images but in five

cases red light images showed degeneration of the pigment epithelium at the site of the lesion

Results

- 71 -

and in one of these cases the pigment-ring lesion as a black area. Congenital hypertrophy of

the retinal pigment epithelium appeared as a variably pigmented, well-delineated, flat lesion

in digital colour, red-free and red light images. A choroidal haemorrhage appeared in digital

colour images as localized dark red or brownish elevation, over which the retinal vessels

course. Red-free images showed the haemorrhage as black with grey or whitish patches due to

changes of the overlying retina. Red light images revealed grey and black areas in the well

delineated lesion. Absorption of the haemorrhage was seen during a follow-up of six months,

and a yellowish white atrophic scar was seen in colour images one year later.

Of the 110 eyes examined by comparative use of colour, red-free and red light imaging there

was exact agreement in 108 cases (98.2%; 95% CI 93.6 to 99.8) between the digital imaging

and the reference standard in assessment of diagnosis of cases referred due to suspected

choroidal melanoma. Digital imaging showed one undercall (0.9%) of a small (4.5 x 5.5 mm

at basal diameter and 2 mm in thickness) amelanotic choroidal melanoma as an amelanotic

choroidal nevus, and one overcall (0.9%) of a large (10 mm at largest basal diameter), heavily

pigmented peripheral melanotic choroidal nevus as a small choroidal melanoma. Digital

imaging showed 85.7% (95% CI 42.1 to 99.6) sensitivity and 99.0% (95% CI 94.7 to 99.9)

specificity in comparison with the reference standard for detection of small choroidal

melanoma from all choroidal tumors suspected to be choroidal melanoma (n = 110); and

85.7% (95% CI 42.1 to 99.6) sensitivity, 98.9% (95% CI 93.8 to 99.9) specificity, 1.1%

undercalls, and 97.9% (95% CI 92.5 to 99.7) exact agreement for detecting it from all

melanocytic choroidal tumors (n=94). Direct comparison between combined use of digital

colour, red-free and red light imaging and the reference standard showed excellent agreement

in detecting small choroidal melanoma from all suspected choroidal lesions (κ 0.847; 95% CI

0.639 to 1.0) and from all melanocytic choroidal tumors (κ 0.846; 95% CI 0.636 to 1.0).

Results

- 72 -

The usefulness of the subtraction method to reveal growth of the choroidal tumour was

evaluated in 94 cases including 78 with melanotic and 16 amelanotic choroidal tumors. In 60

cases the method could not be used because in 46 cases including 39 with melanotic nevi, 4

amelanotic nevi, and 3 melanomas the colour images of the lesion were only available from

one visit, and in 14 cases including 7 with melanotic nevi and 7 amelanotic nevi the lesion

was too diffusely bordered or peripheral. In 34 cases the subtraction method could be used

with the follow-up (time between the used earlier and later digital images) varying between 6

and 68 (mean 28 ± 19) months. In 30 cases, including 26 with melanotic nevi and 4

amelanotic nevi the subtraction method was useful but did not show growth of the lesion. Of

the seven cases with small choroidal melanomas the subtraction method could not be used in

three cases including two melanotic and one amelanotic choroidal melanoma, because the

patient was referred for treatment after the first visit. In four melanocytic choroidal tumours

the subtraction method revealed growth of the lesion during a retrospectively determined

minimum follow-up of 9 to 36 (mean 23 ± 11) months, and the diagnosis was reclassified as

choroidal melanoma.

5.3 Digital infrared transillumination imaging of normal iris (III)

In the normal iris digital IRT images showed the sphincter muscle as a dark circular band in

the pupillary zone of the iris in all eyes. The posterior surface of the iris showed the delicate

radial contraction folds of Schwalbe in the pupillary zone in 58 of the 60 examined normal

eyes (96.7%). The more prominent radial structural folds of Schwalbe were seen in the ciliary

zone in all eyes. The circular contraction furrows were seen in 55 eyes (91.7%). The anterior

iris stroma including the vessel layer and the anterior surface of the iris was poorly outlined.

In 14 eyes (23.3%) with translimbal or corneal illumination some parts of the collarette or

Results

- 73 -

radial iris vessels were seen as light bands. At the side of the illuminating probe a reflex of the

temporal equatorial border of the lens was outlined as a dark circular band in 56 eyes

(93.3 %), whereas in four eyes with translimbal or corneal illumination the equatorial border

of the lens was not discernible. Neither the gender nor the iris colour had any significant

effect on the visualisation of the iris structures and the reflex of the temporal equatorial border

of the lens with IRT imaging. If the illuminating probe was located posteriorly near the

equator, the ciliary body could also be visualised (Fig 4).

Biomicroscopy and digital colour images of the anterior eye clearly showed outlined

pigmented iris patches or iris nevi in 12 eyes (20%). These were not discernible in the infrared

transillumination images of six eyes, whereas more dense or thicker iris nevi were visualised

as dark areas in infrared transillumination images of six eyes. Infrared transillumination

images showed an iris cyst in one (1.7%) of the 60 examined normal eyes. In two eyes

(3.3%), including the right eye of the 30-year-old man with a history of a blunt eye injury 18

years earlier, the iris looked somewhat atrophic.

5.4 Iris atrophy, serous detachment of the ciliary body and ocular hypotony in

chronic phase of VKH disease (IV)

A 52-year-old woman, who had been treated for chronic bilateral uveitis with prednisolone

eye drops for two years, and operated on for complicated cataracts of both eyes one year

earlier, was referred to the Department of Ophthalmology, Turku University Central Hospital

on August 21, 2000. Visual acuity was 0.04 in the right eye and 0.6 in the left, the IOP was

0-1 mmHg in the right eye and 2-3 mmHg in the left. The iris was greyish in both eyes

showing 3+ aqueous flare and 1+ cells. Digital IRT imaging showed for both eyes small and

large patches of atrophy both in the pupillary and ciliary zones of the iris and gathering of

Results

- 74 -

pigment in the inner half of the ciliary zone, and detachment of the ciliary body. In both eyes

oedema of the optic disc and a pale fundus occurred. The macular area was oedematous,

showing chorioretinal folds. In the posterior pole between the superior and inferior temporal

vessels the sensory retina showed a shallow elevation. In the midperiphery, the fundus

showed mottled hyperpigmentation and small patches of depigmentation. Fluorescein

angiography showed chorioretinal folds in the posterior fundus, accumulation of fluorescein

in the subretinal space, some diffusely scattered dots of hyperfluorescence due to window

defects of retinal pigment epithelium, and diffuse fluorescence of the disc in the late phase.

Ultrasonography showed shallow serous detachment of the retina in the posterior pole.

Vitiligo in both forearms, upper arms, thighs, legs, and on the left wrist, sensoryneural

hearing-loss in the audiogram of both ears, pleocytosis in the cerebrospinal fluid, and the

HLA phenotypes confirmed the diagnosis of VKH disease.

The treatment was started using 80 mg prednisolone a day, topical prednisolone eye drops

every hour daily to both eyes and prednisolone ointment at night, plus scopolamin eye drops

twice daily. The eye condition did not respond to the prednisolone treatment. Cyclosporin A

with a daily dose of 250 mg (3 mg/kg) was added to the treatment and

continued for two years.

On May 16, 2002 corrected visual acuity was 0.6 in both eyes. There was 1+ aqueous flare

but no cells in both eyes. The IOP was 4 mmHg in the right eye and 5 mmHg in the left. The

iris was greyish blue in both eyes. Conventional transpupillary transillumination with white

light only showed minute patchy atrophy of the pigment epithelium in the pupillary zone.

Digital infrared transillumination imaging showed extensive atrophy of the iris stroma and

occasional small pigment clumps both in the pupillary and ciliary zones of the iris, and

Results

- 75 -

detachment of the ciliary body in both eyes. The treatment did not normalise bilateral shallow

retinal detachment of the posterior pole, serous detachment of the ciliary body, and severe

ocular hypotony.

5.5 Infrared transillumination imaging and fluorescein angiography of iris nevus

and melanoma (V)

5.5.1 Infrared transillumination imaging.

Infrared translucency (37% in iris nevi and 40% in iris melanomas) was not a differentiating

factor between an iris nevus and melanoma (p = 0.53). IRT images showed the iris nevi as

translucent and ill defined (8%), partly translucent with a lighter and darker area at the site of

the nevus (29%), or as black and mostly well delineated area not transmitting IR light (63%).

Table 7. Infrared transillumination and fluorescein angiographic findings in 58 consecutive

melanocytic iris tumours

Variable Categories Nevus Melanoma P-value (N = 53) (N = 5)

Infrared transillumination Translucent 4 (8%) 2 (40%) Partly translucent 14 (29%) 0 0.53 Opaque 31 (63%) 3 (60%) Not available 4

Masking in iris angiography None 7 (18%) 2 (40%) Incomplete 8 (20%) 0 0.69 Complete 24 (62%) 3 (60%) Not available 14

Filling pattern of tumour Normal 31 (79%) 3 (60%) Disorganised vessels 8 (21%) 2 (40%) 0.32 Not available 14

Fluorescein leakage from None 26 (67%) 3 (60%) tumour vessels Early moderate 5 (13%) 0 1.0 Late moderate 5 (13%) 0 Late gross 3 (8%) 2 (40%) Not available 14

Results

- 76 -

The iris melanoma was translucent to IR light in two cases, showing in IRT images as a light

area at the site of the tumour. In three cases the iris melanoma did not transmit IR light and it

appeared in IRT images as a black area, which extended in one case partly into the ciliary

body (Table 7).

5.5.2 Fluorescein angiographic findings

Complete masking of fluorescence (62% in iris nevi and 60% in iris melanomas, p = 0.69),

presence of disorganised vessels (21% in iris nevi and 40% in iris melanomas, p = 0.32), and

fluorescein leakage (34% in iris nevi and 40% in iris melanomas, p = 1.0) at the site of the

lesion did not differentiate an iris nevus froma a melanoma (Table 7).

Discussion

- 77 -

6. DISCUSSION

6.1 Sensitivity and speficifity of digital retinal images in grading diabetic

retinopathy (I)

6.1.1 Digital red-free and colour imaging in grading diabetic retinopathy (I)

Red-free photographs showed well the superficial layers of the retina and retinal

vessels.47,79 We showed that digital red-free imaging had the best sensitivity (97,7 %;

95%CI 95.8-99.7) and specificity of 98,8%; 95%CI 96.7-100 for diagnosis of mild

NPDR. We observed that microvascular changes of DR, including MAs and IRMA,

were most clearly visualised with digital red-free imaging, which is in good agreement

with earlier observations based on red-free photographs and red-free black and white

images.100,104,243 Overcalls in the digital red-free imaging versus the reference standard

support earlier findings that red-free wide-field digital imaging was more sensitive for

DR screening than mydriatic ophthalmoscopy, the currently accepted screening method,

and more sensitive than the 35 mm colour slides.104,141

We showed that digital retinal colour imaging (Topcon TRC 50 IA and Canon CR6-

45NM) had a sensitivity of 88.9-94.0% and a specificity of 99.0-100% for diagnosis of

mild NPDR. Thus both digital colour and red-free retinal imaging surpass the criteria of

80% sensitivity and 95% specificity set for the acception of a screening method for

sight-threatening DR.24,227 Our results only showed 1.2-1.6% ungradeable digital retinal

colour images and 1.3% digital red-free images. These figures are lower

Discussion

- 78 -

than the 5% failure rate for any photographic method acceptance for screening of

DR.24,227 Kappa values showed very good agreement between colour and red-free

imaging.

We used single (still) images produced by a commercial version of the handheld digital

colour video camera (MediTell) in the grading of DR (I). The images were examined in

a multicenter study by three examiners in a randomised and masked manner. The

handheld digital video camera showed a sensitivity of 6.9% (95% CI 2.3 – 11.5) and

ungradeable images represented 92.3 % (I). The quality of the single (still) images of

the ocular fundus produced by the handheld video camera (MediTell) (I) was worse

than that of the continuous fundus colour videophotography. 192 Still images were

obscure and showed only occasionally haemorrhages and hard exudates, but could not

demonstrate MAs, IRMA or neovascularization. This can be explained by the fact that

although the pixel count reflecting the resolution of the system was better in MediTell

(768 x 576 pixels) than that of most basic video systems, it was only slightly more than

one-third of the digital Canon CR 6-45NM fundus camera (2160 x 1440 pixels).203

Furthermore, the imaging field of MediTell was about three disc diameters (15º) and

imaging was possible only at about 20º peripheral to the fovea. These attributes explain

the high percentage of ungradeable images and the low sensitivity of MediTell for

detection of at least mild NPDR. These figures do not reach the criteria for DR

screening.24,227 Similarly, in an earlier study, the handheld Nidek NM-100 fundus

camera was not suitable for diagnosis of DR.253 The MediTell camera should be further

developed by improving the focusing qualities, use of a head-rest, increasing the

imaging field, and by providing the possibility of red-free imaging.

Discussion

- 79 -

6.1.2 The effect of the size of the imaging field in grading diabetic retinopathy (I)

The size of the imaging field had an effect on the sensitivity of the imaging modality.

The digital retinal colour and red-free imaging with the 50º field Topcon TRC 50 IA

camera showed the best sensitivity followed by the 45º field Canon CR6-45NM camera.

The sensitivity was significantly worse when using the 15º field MediTell camera which

made digital imaging possible at only about 20º from the central fundus but the poor

sensitivity and specificity reflected not only the size of the imaging field but also the

camera’s inferior picture quality. An earlier study reported that in photographic

screening a wide field is of benefit in detection of neovascularization.245 Two-field 45º

photography disclosed 77% of neovascularizations detected by two-field 60º

photography.244

6.1.3 The role of screeners in grading diabetic retinopathy

Ideally, screening of DR should be done by a dedicated ophthalmologist.184 We found

no significant differences in the sensitivity figures of different imaging modalities for

screening DR between the specially trained bachelor of medicine and the

ophthalmologists. Similarly, in two earlier studies the quality of the screening results

was not related to the medical specialty of the screeners.104,140 These results support the

suggestion that other health care providers can screen,provided that they are thoroughly

trained and may have a dedicated ophthalmologist as consultant.184

6.1.4 Screening of diabetic retinopathy

It is important to establish effective screening programs for detection of DR, which in

its early easily treatable form is asymptomatic and is a common cause of preventable

blindness.140,224 A general consensus exists concerning the cost-effectiveness of

Discussion

- 80 -

screening DR.21,119,140,224 Our results (I) showed that digital red-free and colour imaging

of the fundus provided sensitivity, specificity, and exact agreement good enough to be

recommended for screening of DR. They also support earlier findings that in

photographic screening a wide field is also of benefit in detection of PDR.224,242,244,245

Photography allows conservation of resources when a population is screened for DR

instead of forcing an ophthalmologist to check the eyes of every patient with a

biomicroscope, he can just examine photographs taken by a specially trained nurse.

Digital imaging allows immediate, magnified images that can easily be analysed,

enhanced, archived, printed or transferred to remote computers thus allowing

telemedicine and telecare of DR.140

6.2 Digital imaging in differential diagnosis of small choroidal melanoma (III)

6.2.1 Comparative use of digital colour, red-free light and red light imaging

Comparative use of digital colour, red-free, and red light imaging had a 85.7% (95% CI

42.1 - 99.6) sensitivity and 99.0% (95%CI 94.7 - 99.9) specificity for diagnosis of

choroidal melanoma. It surpasses the 80% sensitivity and 95% specificity levels

considered acceptable for a photography screening method for sight-threatening eye

disease.226 No (0%) ungradeable digital images occurred which is lower than the 5%

failure rate considered acceptable for any photographic method for screening of the eye

disease.226 We found a 98.2% (95% CI 93.6 - 99.8) exact agreement between the use of

digital images and the reference standard in differentiation of small choroidal melanoma

from pseudomelanomas. Direct comparison between the use of digital images and the

reference standard showed excellent agreement in detecting choroidal melanoma from

suspect choroidal lesions (κ 0.847; 95%CI 0.639 - 1.0).

Discussion

- 81 -

The clinical diagnosis of choroidal melanoma by an ophthalmic oncologist is accurate

in over 98% of instances when the tumour size is medium or large, and the COMS

misdiagnosis rate of 0.48% for medium and large-sized choroidal melanomas is the

lowest ever reported.36 Digital colour images may be regarded as a surrogate for

ophthalmic physical examination by fundus biomicroscopy. Together with the red-free

retinal imaging and the red light choroidal imaging it yielded a high diagnostic accuracy

including 85.7% sensitivity, 99.0% specificity, and 98.2% exact agreement versus

reference standard in diagnosing small choroidal melanoma and different

pseudomelanomas.

We found that one undercall and one overcall in the use of digital imaging versus

reference standard in the diagnosis of choroidal nevus and melanoma indicates that by

using digital images only it is difficult to differentiate between a suspicious choroidal

nevus and a small choroidal melanoma always correctly. Standard clinical information,

including demographic data, medical history, presenting symptoms, mydriatic contact

lens funduscopy, comparative use of digital colour, red-free, and red light imaging,

FAG, B scan ultrasonography and if needed, ICG angiography, and orbital CT scan and

MRI yield a higher diagnostic accuracy than the digital colour, red-free, and red light

imaging alone. Digital non-stereo imaging of the ocular fundus only gives

two-dimensional images of the lesion. Ultrasonography is a particularly useful clinical

method giving accurate measures of the thickness of the lesion. In addition, scanning

laser ophthalmoscopical ICG angiography, OCT, and diagnostic transvitreal fine-needle

aspiration biopsy may have value for special questions in the evaluation of small

choroidal tumours in departments of ophthalmic oncology.3,5,161 Digital images can be

Discussion

- 82 -

delivered for diagnostic purposes not only via the image net to different departments of

the same eye center but can be transmitted by electronic mail via the Internet to distant

experts of ocular oncology for second opinion of the fundus lesion.

6.2.2 Subtraction method

Suspicious choroidal nevi require follow-up to detect growth and to clinically verify the

diagnosis of choroidal melanoma.26 When growth is documented, treatment is generally

recommended although documented growth is not an unequivocal indicator of

melanoma for small melanocytic tumours.5,66,146 We described a new subtraction

method to demonstrate early growth of the melanotic choroidal tumour (III). Four

melanocytic choroidal tumours (4.3 % of all 94 melanocytic choroidal tumors, and 11.8

% of those studied by subtraction method) showed growth during a minimum follow-up

of 9 to 36 (mean 23 ± 11) months. This finding is in good agreement with an earlier

study showing growth in 4.5% of choroidal nevi, 14.2% of suspicious nevi, 50.1% of

dormant melanomas, and 86.1% of active melanomas during 5 years of follow-up.8

Because there were also 5.8 to 16% metastatic and death rates for patients with growing

small choroidal melanomas during a 5-year follow-up, a need for early diagnosis and

treatment of those tumours exists.26,51,67,68 The subtraction method we described may be

a valuable adjunct to comprehensive physical examination for early diagnosis of small

choroidal melanoma in cases which are situated in the area of fundus which can be seen

in digital images.

6.2.3 Epidemiological aspects of choroidal melanoma

We found a mean annual incidence of choroidal melanoma of 0.80 per 100,000

population in the Hospital District of Southwest Finland during 1987-2003. This is in

Discussion

- 83 -

good agreement with the incidence of uveal melanoma varying from 0.6 to 1.2 per

100,000 population during 1960-1998 in Sweden.19 The annual number of new cases of

uveal melanoma has showed a slight increase with time, which may be accounted for by

the increase in the old age groups of the population.175 Because the incidence of uveal

melanoma is low, it is not feasible to mass screen even the age group which would be at

highest risk of developing it.64 As a rare disease, a small choroidal melanoma may

remain unnoticed, and in the office of a general ophthalmologist the misdiagnosis rate

of ocular melanoma was 29% in Finland and 37% in San Francisco, California.23,68 On

the other hand, in our study 75% of the patients referred to a tertiary center with the

diagnosis of choroidal melanoma proved to have a pseudomelanoma. The results

suggest that in the remitting eye centers combined use of digital colour, red-free, and

red light imaging of the ocular fundus may help in the diagnosis and differential

diagnosis of small choroidal melanoma. It is an easy, rapid and useful method for the

diagnosis, differential diagnosis, and follow-up of choroidal melanocytic tumors in

every eye center with a digital imaging system.

6.3 Digital infrared transillumination imaging of the iris (III, IV, V)

Progress from the analog stereophotography of the anterior eye to digital imaging

suggested the development of a new method for digital IRT imaging of the iris. In our

study the illumination taken from the Zeiss slitlamp 6V 30W incandescent lamp was

guided through a Kodak Wratten filter 87 transmitting only IR light (Table 1, Fig. 2). In

the 1.9 m long light fiber optic cable the attenuation was only about 0.008-0.04 dB,

because in modern optic fibers the attenuations is 4-20dB/km.98

Discussion

- 84 -

A major part of the optical radiation is backscattered by the sclera which accounts for

the white colour of the sclera.73 The strong light scattering within the sclera is mainly

caused by the heterogenous distribution and the variable diameter of the collagen

fibrils.73,106 Another reason is the difference of the refractive indices between the

collagen fibrils (n=1.47) and the ground substance surrounding the fibrils (n=1.36).73

Fine at al. observed that 0.07 – 0.09 of the 632 nm light was transmitted by the sclera.73

With increasing wavelength the scattering in scleral tissue and its optical density

became weaker (Fig. 5).73 Vogel et al. observed that the scleral transmission was only

6% at 442 nm but increased to 35% at 804 nm and to 53% at 1064 nm. 236 Thus, in OCT

anterior segment imaging the IR light of 830 nm showed limited penetration depth

through the highly backscattering sclera, but the 1310 nm IR light was useful in

reducing backscattering of the sclera and in improving imaging of the iris and ciliary

body.106

Fig. 5. Dependence of relative optical density of sclera on the wavelength of the incident radiation.

Experimental results of three different sclera (A, B, C).73

Discussion

- 85 -

Fig. 6. Measured refractive index vs. wavelength and dispersion curves for human aqueous and vitreous

humor.205

Vogel et al. observed that the scleral absorption was 40% at 442 nm but only 6% at 804

nm and 1064 nm. 236 Fiber contact increases transmission, with a factor of 3.5 at 442

nm, of 2.0 at 804 nm, and 1.5 at 1064 nm.236 The biological tissue effect of the laser

light depends not only on the total transmission of the sclera, but also on the directional

distribution of the transmitted light. 236 The pigment epithelium of the ciliary body and

choroid drastically reduces the red light transmitted by the sclera.73 The penetration of

the sclera by red light and IR rays has been used in the transcleral

cyclophotocoagulation treatment of uncontrolled glaucoma.127,176 The IR Nd:YAG laser

is most widely used for this purpose. Transcleral red-laser cyclophotocoagulation

Discussion

- 86 -

combined with either 647 nm krypton or a 670 nm diode laser may also be used

although they penetrate the sclera less well than the IR Nd:YAG laser but are better

absorbed by the pigment epithelium in the ciliary body.127,176

Sardar et al. studied the optical properties of human aqueous humor and vitreous.205 The

measured refractive indices at different wavelengths and dispersion curves are shown in

Fig. 6. Table 8 shows their results on measured diffuse reflectance, total transmittance

and index of refraction values for human aqueous humor and vitreous at 980 nm and

1310 nm.

Table 8. Measured diffuse reflectance (DR), total transmittance (TT) and index of

refraction (n) of human aqueous humor and and vitreous at 980 nm and 1310 nm.18

Wavelength (nm) Sample DR TT n

980 Aqueous 0.0041 0.8416 1.33 humor

980 Vitreous 0.0030 0.8266 1.34

1310 Aqueous 0.0038 0.7671 1.34 humor

1310 Vitreous 0.0018 0.668 1.34

Berg and Spekreijse studied the near IR light absorption in the human eye media and

represented for cornea, aqueous, lens and vitreous the transmittance spectra replotted

here from literature (Fig 7).18 They concluded that in the near IR region the light losses

in the eye media are best estimated with the absorption coefficient for pure water.18

Discussion

- 87 -

Fig. 7. Transmittance spectra for the eye media replotted from literature.18

Our digital IRT findings are in good agreement with the above transmittance and

absorption calculations. In the clinical digital IRT images of the iris the IR light coming

through the pupil appeared somewhat weaker than that going through the sclera (Fig. 3

and 4 pg 56).

The digital IRT imaging of the iris showed the stroma and posterior surface of the iris

including the sphincter muscle and the radial contraction folds of Schwalbe in the

pupillary zone, and the radial structural folds of Schwalbe in the ciliary zone in the same

way as with the IRT stereophotography of the iris.195,198 The in vivo findings of the

structural pattern of the posterior iris surface we observed are in good agreement with

those observed with light and scanning electron microscopical examinations.52,58

Discussion

- 88 -

Circular contraction furrows were somewhat better seen in IRT stereophotographs than

in the digital IRT images of the iris.199 On the other hand, digital IRT images showed

the reflex of the temporal equatorial border of the lens in 93% of the eyes, whereas this

finding was not reported in IRT stereophotographs.199 These differences may be

explained on the basis that infrared transillumination stereophotographs were taken with

a standard Zeiss stereo slit-lamp, camera whereas digital IRT images were taken with a

digital Topcon retinal camera.

Digital colour imaging showed the anterior surface of the iris and the IRT imaging the

structural pattern of the stroma and posterior surface of the iris. The iris nevi seen in

colour photographs were not clearly discernible in IRT stereophotographs but a thick,

elevated pigmented patch did not transmit infrared light.199 Similarly, we saw more

dense or thicker nevi as dark areas in digital IRT images of the iris.

The Finns are one of the fairest populations in the world. In our study, 73.3% of the

eyes were classified to grade 1 and 16 eyes were classified to grades 2-5 of Seddon et

al.209 The results of this study showed that neither the colour of the iris nor gender had

any significant effect on the visualization of the structures of the stroma and posterior

surface of the iris or the temporal equatorial border of the lens in digital IRT images of

the iris.

Digital IRT imaging is convenient for the patient, easy to repeat, and it does not require

any stain injection or darkening of the examination room. Digital imaging systems

provide immediate, magnified images that can easily be analysed, enhanced, archived,

printed or transferred electronically to remote computers. Adaptation of the light and

Discussion

- 89 -

contrast of the digital images of the iris helped to visualise different changes of the iris

structure and pigmentation. Because no film was used, small foreign bodies and other

artefacts of the film development stage were avoided.

By using the transcleral transillumination with bright light the photographs of the iris

showed widespread pseudoluminance of the exceptionally fair normal iris and

generalised the peripheral iris transluminance in 45% of eyes with pseudoexfoliation

syndrome and in 18% of the eyes in the control group (p < 0.01). All the eyes in that

study were blue to bluish-gray in colour.183 In transcleral transillumination photography

the very bright white light directed through the temporal sclera as posteriorly as possible

may cause greater translucency of the iris than the reflection of the light beam of the

slit-lamp in the transpupillary transillumination of the iris. We found that (III-V) digital

IRT imaging of the iris showed the structural changes of the posterior surface and

stroma of the iris more evenly and clearly and the iris colour did not have any

significant effect on the visualization of the iris structures with IRT images. Different

findings in the transcleral transillumination photography of the iris and digital IRT

imaging of the iris depend on the different wavelengths used. Transcleral

transillumination photography registered the atrophic areas of the iris seen by using

visible light and ASA 400 film, where the panchromatic emulsion was not sensitive to

radiation beyond the visible red, or about 700 nm.62 In the digital IRT imaging we used

the Kodak Wratten filter 87 cutting away all radiation below 740 nm (Fig. 2 pg 55).

Anterior segment OCT may be used to study cross-sections of the iris. It is comfortable

for the patient due to its noncontact nature. The IRT imaging of the iris shows changes

Discussion

- 90 -

of the posterior surface and stroma of the whole iris. Thus, digital IRT imaging of the

iris and anterior segment OCT may be used to supplement each other.

6.4 Infrared transillumination findings of the iris and ciliary body in chronic

phase of Vogt-Koyanagi Harada disease (IV)

We showed the usefulness of digital IRT imaging for examination of the iris and ciliary

body in VKH disease. In the chronic phase of VKH disease both biomicroscopy and

conventional digital colour imaging showed the iris rather normal and the conventional

transpupillary transillumination technique by using the white light only showed minute

patches of atrophy of the pigment epithelium in the pupillary zone of the iris even four

years after beginning of the uveitic phase. Using IRT imaging two years after the

beginning of the uveitic phase of VKH disease we showed small and large patches of

iris atrophy in the pupillary and ciliary zones in both eyes, and a large gathering of

pigment to the inner half of the ciliary zone. A further two years later in the chronic

phase of VKH disease an extensive atrophy of the iris stroma existed, but most of the

pigment gathering had disappeared, and only small pigment clumps were seen in the

pupillary and ciliary zones of the iris in both eyes. In addition, IRT images showed

detachment of the ciliary body in both eyes. In the iris stroma IRT imaging reveals the

nodules and atrophic patches more clearly than other clinical methods. Thus, it can be

used in the diagnosis and follow-up not only in VKH disease, but also e.g. in sarcoid

uveitis and in Fuchs’ heterochromic cyclitis.121,200

The present case helps us to understand the morphological changes of the iris stroma in

VKH disease. In the uveitic phase the iris is thickened by diffuse infiltration of

lymphocytes, macrophages, and epitheloid cells.115,168 Iris nodules may be noted on the

Discussion

- 91 -

pupillary margin, as well as in the iris stroma in 5 % of the cases in the uveitic phase

and in 29 % of the cases in the chronic phase of VKH disease.168,189 These nodules

transmit IR light and may be seen in the uveitic phase as round light patches in IRT

images of the iris in the same manner as reported in sarcoid uveitis.121 In our study, two

years after the beginning of the uveitic phase of VKH disease, the nodules had become

atrophic and appeared as small and large patches of atrophy in IRT images. Melanocyte

destruction was seen as a large gathering of pigment in the inner half of the ciliary zone.

Two years later in the chronic phase of VKH disease, the iris showed extensive atrophy

and only occasional small pigment clumps. The chronic phase of VKH disease has been

characterised by the depigmentation of the fundus (sunset glow fundus) and limbus

(Sugiura sign).116,178 The Sugiura sign does not occur in caucasian patients with VKH

disease.168,189 We suggest that characteristic findings of the chronic phase of VKH

disease may together with the sunset glow fundus also include the depigmentation and

atrophy of the iris stroma.

Cataracts may develop in 10.5-40 % of patients with VKH disease.158,159,178,189 In VKH

disease a cataract operation can be recommended only in eyes where the inflammation

has been controlled for at least two to three months with high dose systemic

corticosteroids.159 The present case had already developed cataract two years prior, and

was operated on in a private eye center during the active uveitic phase without any

preoperative control of inflammation one year before admittance to the Turku

University Central Hospital. In Finland VKH disease is very rare and only one case has

previously been reported.171,195

Discussion

- 92 -

Glaucoma may occur in VKH disease in 6-45 % of the cases.81,159,178,189 Of these 56 %

had open angle glaucoma and 44 % angle closure glaucoma.81 On the other hand, severe

inflammation of the anterior uvea may cause oedema of the ciliary body and swelling of

the ciliary processes early in VKH disease, and may occasionally lead to transient

hypotony in the uveitic phase of the disease.126,168 We are not, however, aware of any

reported cases of chronic hypotony in VKH disease. Our patient already had bilateral

hypotony before the cataract surgery and over one year before the peroral corticosteroid

and cyclosporine treatment was started. Ultrasound biomicroscopy showed in the

uveitic phase of VKH disease, the transient accumulation of exudation between the

ciliary body and sclera.123 After treatment with systemic corticosteroids, the supraciliary

exudation disappears. If steroid therapy is stopped too soon, recurrence or progression

of disease may ensue and ciliochoroidal detachments may persist.162 The present patient

remained without any systemic corticosteroid treatment for two years after the

beginning of the uveitic phase of the disease. In addition, during this time she

underwent cataract extraction and intraocular lens implantation in both eyes which may

have further deteriorated the situation. The IRT images showed a persistent detachment

of the ciliary body in both eyes two and four years after the beginning of the uveitic

phase. Atrophy of the ciliary processes and cyclitis as part of the severe chronic anterior

uveitis,115 and detachment of the ciliary body may explain the ocular hypotony in this

case. Future studies with larger material may confirm the findings seen in our case.

Evaluation of the ocular fundus of our patient with chronic VKH disease showed

macular oedema and subtle chorioretinal folds in the posterior pole, and patches of

depigmentation and mottled hyperpigmentation causing a “blond” sunset glow

fundus.168 Macular oedema has been reported in eyes with a more protracted chronic

Discussion

- 93 -

VKH disease.190 In our case with aggressive therapy for two years the shallow

detachment of the macular area did not reattach. Obviously poor function of the pigment

epithelium, chorioretinal folds, and ocular hypotony precluded normal reattachment of

the sensory retina in the posterior pole.

In VKH disease it is important to suppress the initial intraocular inflammation with

early and aggressive use of systemic corticosteroids, followed by their slow tapering

over three to six months.159,168 This treatment may shorten the duration of the disease

and prevent progression into the chronic stage. Steroidresistant recurrences may only

respond to cytotoxic/immunosuppressive agents.159,168 In the present case the patient

was admitted to the University Hospital in the chronic phase. The patient was treated for

two years with peroral and topical prednisolone and with oral cyclosporine. During the

treatment the aqueous cells disappeared but the treatment did not have any improving

effect on the hypotony or iris atrophy, and the ocular fundus still showed oedema of the

optic disc and macular area in both eyes.

6.5 Infrared transillumination imaging and fluorescein angiography in differential

diagnosis of iris nevus and melanoma (V)

This is the first study to statistically evaluate the differentiation of iris nevus and

melanoma by using both IRT imaging and FAG of the iris. IRT imaging showed that

the lesion was translucent in 40 % of iris melanomas and in 8 % of iris nevi, partly

translucent in 29 % of iris nevi, and opaque to IR light in 60 % of iris melanomas and in

63 % of iris nevi. Thus IRT imaging cannot be used to differentiate iris nevus and

melanoma (p = 0.53). IRT images are useful in defining the area of tumour

Discussion

- 94 -

involvement. It can identify iris atrophy, cysts, foreign bodies, and granulomas,196 and

in this way it also contributes to the differential diagnosis of iris melanoma.

In an earlier study complete masking in iris angiograms at the site of the lesion

occurred in none of the four iris melanomas but in 4/9 iris nevi,41 a difference which

can be calculated to be not statistically significant (p = 0.10). In our study, complete

masking occurred in 62 % of iris nevi and in 3/5 iris melanomas, indicating that

complete masking in iris angiograms cannot be used to differentiate between iris nevus

and melanoma (p = 0.69).

Observed growth of a melanocytic iris lesion is a predicting factor of malignancy.97,229

A clinical follow-up study showed that vascularity of the lesion was unassociated with

the growth of the melanocytic iris tumour (p = 0.42).229 Fluorescein angiograms of the

iris showed irregularly arranged vessels both in five histologically confirmed nevi and

in four melanomas.48 Histological studies showed prominent superficial vasculature

only in 10 - 20 % of iris melanomas.124 Our results are in good agreement with these

earlier studies. In our study, disorganised vessels were seen in 21% of 53 iris nevi and

in two of five iris melanomas, suggesting that disorganised vessels were not diagnostic

for malignancy (p = 0.32). In an earlier study, fluorescein leakage occurred in all four

iris melanomas and in 5/9 iris nevi,41 a difference which can be calculated to be

statistically nearly significant (p = 0.07). In our larger study, however, fluorescein

leakage was not a significant predicting factor for iris melanoma (p = 1.0). Our results

confirm the view that fluorescein angiograms provide little useful diagnostic aid to

distinguish between benign iris nevi and melanomas.6,7,124

Discussion

- 95 -

Our results (V) suggest, that suspicious iris nevi should be observed carefully over time

by using slit-lamp photography or digital colour imaging to evaluate growth regardless

of their IRT imaging and fluorescein angiographic characteristics. When growth occurs,

local excision of the tumour is usually indicated. Four of the eight melanocytic iris

tumours, however, which were excised due to growth were histologically proved nevi.

This is in good agreement with earlier findings showing that growth does not always

confirm malignancy as both benign and malignant tumours including iris nevi and

spindle A melanomas can grow and invade the trabecular meshwork.59,215 Nevertheless

growth is still the single most consistent indicator of malignancy and the periodic

measurement of tumour size is a standard procedure. IRT images and fluorescein

angiograms cannot be used to determine if the melanocytic iris tumour is benign or

malignant.

6.6 Teleophthalmology and future directions

The invention of colour CCD cameras has enabled tertiary referral centres, university

eye clinics and other facilities to capture and store digital pictures on a computer.211 The

price of the technology remains prohibitive to smaller facilities. The high equipment

costs of the camera and image acquisition system are compensated by the minimal cost

of additional images, possibility for telecommunication, and no need for films and their

development to paper prints, archiving of photographs, and time consuming manual

image evaluation. Telemedical projects encourage central planning and the collecting of

patient records from primary care units so that they are readily available for use by

specialists in the referral centre.169 Technical failures happen less often when using

digital imaging instead of film.210

Discussion

- 96 -

Converting an analog signal to digital form causes a loss of information. Therefore the

theoretical maximum resolution of digital imaging is always worse than that of film.

Today in experimental studies the images of the future digital fundus cameras have a

spatial resolution of 9μm/pixel for a 30 degree field, about twice the SLO resolution. 251

In clinical work, the digital image is displayed on a large computer screen allowing full

resolution whereas film edges reduce the effective resolution of film-based photographs.

Duplicates of digital images can be produced with no loss of resolution. Magnifying an

image e.g. for laser guidance is enhanced with digital images.

The capacity of computer technology to calculate has gone up exponentially and the

basic technological know-how exists to create many tools that have not yet found wide-

spread use in ophthalmology. Pioneering work in testing and developing new equipment

and software applications is required. The price of complicated medical equipment

gives importance to comparative testing between equipment of different companies.

The advances of digital imaging systems and telecommunication technology have made

digital imaging the new standard in the ophthalmic community.251 The ICG angiography

and OCT can only be obtained using digital imaging. It is likely that film-based imaging

will vanish from the profession within the next decade.251

Summary and Conclusions

- 97 -

7. SUMMARY AND CONCLUSIONS

This study was initiated to develop and assess new digital photographic methods for the

diagnosis and follow-up of ocular diseases.

1. A randomised and masked multicentre study to assess digital fundus cameras for DR

screening showed that digital 50° red-free imaging (Topcon TRC 50 IA) had 97.7%

(95%CI 95.8 – 99.7) sensitivity, 98.9% (95%CI 96.9 – 100) specificity, and 98.1%

exact agreement with the reference standard for detecting at least mild NPDR. Digital

two-field 50° colour imaging (Topcon TRC 50 IA) showed 94% (95%CI 90.8 – 97.2)

sensitivity, 99.0% (95%CI 96.9 – 100) specificity, and 95.5% exact agreement for

detection of at least mild NPDR when compared with the reference standard. Digital

two-field 45° colour imaging (Canon CR6-45NM) showed 88.9% (95%CI 82.0 – 95.8)

sensitivity, 100% specificity, and 89.3% exact agreement for detection of at least mild

NPDR when compared with the reference standard. Only 1.3% of digital red-free and

1.2 – 1.6% of digital colour images were ungradeable. These digital imaging modalities

surpass the 80% sensitivity, 95% specificity, and 5% failure rate considered acceptable

for any photographic method for screening of DR.24,226 When using the handheld digital

colour video camera (MediTell) the still images taken of the ocular fundus showed only

6.9%; 95%CI 2.3 – 11.5 sensitivity for detection of at least mild NPDR when compared

with the reference standard and ungradeable images represented 92.3%. These figures

did not reach the criteria for DR screening

2. Digital colour, red-free and red light images were evaluated in a randomised and

masked manner and by a new subtraction method for differential diagnosis of small


- 98 -

choroidal melanoma. Comparative use of digital colour, red-free, and red light imaging

had 85.7% (95%CI 42.1 – 99.6) sensitivity, 99% (95%CI 94.7 – 99.9) specificity, and

98.2% (95%CI 93.6 – 99.8) exact agreement versus reference standard in differentiation

of small choroidal melanoma from pseudomelanoma. Excellent agreement occurred

between the comparative use of digital images and the reference standard in detecting

small choroidal melanoma from suspect choroidal lesions (κ 0.847; 95%CI 0.639 – 1.0).

The subtraction method was useful in showing growth in four of 94 melanocytic

tumours (4.3%) during a minimum follow-up of 9-36 (mean ± SD of 23 ± 11) months.

These figures show that comparative use of digital colour, red-free and red light

imaging is a suitable method in differentiation of small choroidal melanoma from

pseudomelanomas. The new subtraction method may reveal early growth of

melanocytic choroidal tumours in cases which occur in the area of the fundus that can

be seen in digital images.

3. In the development of a new digital IRT imaging of the iris and ciliary body the

examination light from a Zeiss photo slit-lamp was passed through a Kodak Wratten

gelatine filter No. 87 transmitting only IR light. A fiber optic cable was used to transmit

the IR light through the lateral wall of the eye. The IR images were captured using a

Topcon TRC 50 IA retinal camera focused on the iris. Digital IRT imaging of a normal

eye showed the sphincter muscle and delicate radial contraction folds of Schwalbe in

the pupillary zone and radial structural folds of Schwalbe and circular contraction

furrows in the ciliary zone of the iris. Digital IRT images also showed the reflex of the

temporal equatorial border of the lens as a dark circular band. Neither the iris colour nor

gender had any significant effect on the visualization of the iris structures with digital


- 99 -

IRT images. Digital IRT images may be used to study changes of the stroma and

posterior surface of the iris in various diseases of the uvea.

4. Ocular examination and follow-up, including digital IRT imaging of the iris and

ciliary body, was done in a 52-year old woman with chronic phase of VKH disease. IRT

imaging showed extensive atrophy of the iris stroma and occasional pigment clumps

both in the pupillary and the ciliary zones of the iris, and detachment of the ciliary body

in both eyes. Treatment did not normalise bilateral shallow retinal detachment of the

posterior pole, serous detachment of the ciliary body, or severe ocular hypotony. We

showed that severe atrophy of the iris stroma, retinal detachment of the posterior pole,

serous detachment of the ciliary body, and ocular hypotony are new findings of chronic

phase of VKH disease.

5. IRT imaging and FAG findings of the iris in 53 eyes with an iris nevus and five eyes

with an iris melanoma showed that IR translucence (37% in iris nevi and 40% in iris

melanomas) was not a differentiating factor between an iris nevus and melanoma

(p = 0.53). Complete masking of fluorescence (62% in iris nevi and 60% in iris

melanomas, p = 0.69), presence of disorganised vessels (21% in iris nevi and 40% in

iris melanomas, p = 0.32), and fluorescein leakage (34% in iris nevi and 40% in iris

melanomas, p = 1.0) at the site of the lesion did not differentiate an iris nevus from a

melanoma. IRT imaging and FAG of the iris are useful in the differential diagnosis of

melanocytic iris tumours but they cannot be used to determine if the lesion is benign or

malignant.


- 100 -

Acknowledgements

- 101 -

8. ACKNOWLEDGEMENTS

I express my deepest gratitude to my supervisors, Professor Tero Kivelä MD, FEBO,

Department of Ophthalmology, University of Helsinki and Professor Matti Saari, MD,

FEBO, Department of Ophthalmology, University of Turku for their guidance in

planning and executing this research project. Their encouraging attitudes and scientific

support have been a persistent force motivating me during each step of this work.

I am grateful to Docent Paula Summanen, MD, FEBO, for kindly consenting to oversee

my research project of screening diabetic retinopathy and for her positive and bright

words and encouragement during this study.

I wish to express my sincere thanks and gratitude to the official reviewers of my thesis,

Docent Eero Aarnisalo, M.D. Department of Ophthalmology, Satakunta Central

Hospital, Pori and Docent Markku Teräsvirta, M.D. FEBO, Department of

Ophthalmology, Kuopio University Central Hospital, Kuopio, for their valuable

criticism and constructive comments, which I highly appreciate.

I wish to thank Professor Juhani Tuominen, Lic in Soc Science, Department of

Biostatistics, University of Turku for statistical advice.

I would like to give thanks to Ophthalmic Photographer Kari Nummelin, BSc,

Department of Ophthalmology, Turku University Central Hospital, without whose

work, assistance and skillful technical expertise it would have been impossible to

conduct research on ophthalmic photography with quality.

Acknowledgements

- 102 -

Finally, I want to thank Erika for always being there.

This study was financially supported by grants from the Finnish Medical Foundation,

Helsinki, Finland and the Instrumentarium Science Foundation, Helsinki, Finland.

References

- 103 -

9. REFERENCES

1. Airaksinen PJ, Mustonen E, Alanko H. Optic disc haemorrhages precede retinal nerve fibre

layer defects in ocular hypertension. Acta Ophthalmol Scand 1981;59:627-41.

2. Airaksinen PJ, Nieminen H, Mustonen E. Retinal nerve fibre layer photography with a wide

angle fundus camera. Acta Ophthalmol Scand 1982;60:362-8.

3. Andersen MVN, Scherfig E, Prause JU. Differential diagnosis of choroidal melanomas and

nevi using scanning laser ophthalmoscopical indocyanine green angiography. Acta

Ophthalmol Scand 1995;73:453-6.

4. Archer DB. Fluorescein angiography. In Krill AE (ed.) Hereditary retinal and choroidal

diseases. Vol. I. Evaluation. New York: Harper & Row, 1972:73-187.

5. Augsburger JJ, Correa ZM, Schneider S, Yassin RS, Robinson-Smith T, Ehya H,

Trichopoulos N. Diagnostic transvitreal fine-needle aspiration biopsy of small melanocytic

choroidal tumors in nevus versus melanoma category. Trans Am Ophthalmol Soc

2002;100:225-34.

6. Augsburger JJ, Damato BE, Bornfeld N. Uveal melanoma. In: Yanoff M, Duker JS, eds.

Ophthalmology, 2nd edn. St Louis: Mosby, 2004:1052–63.

7. Augsburger JJ, Harbour JW, Gonder JR. Uveal nevus. In: Yanoff M, Duker JS, eds.

Ophthalmology, 2nd edn. St Louis: Mosby, 2004:1077–80.

8. Augsburger JJ, Schroeder RP, Territo C, Gamel JW, Shields JA. Clinical parameters

predictive of enlargement of melanocytic choroidal lesions. Br J Ophthalmol 1989;73: 911-

7.

9. Babel J. Syndrome de Vogt-Koyanagi-Harada disease. Schweiz Med Wochenschr

1932;44:1136-40.

10. Von Baeyer A. Ueber eine neue Klasse von Farbstoffen. Ber Deutschen Chem Ges

1871;4:555-8.

11. Baggesen LH. Fluorescence angiography of the iris in diabetics and non-diabetics. Acta


12. Von Bahr G. Farberphotographieren des Augenhintergrundes. Acta Ophthalmol Scand

1938;16:483-4.

References

- 104 -

13. Bandello F, Brancato R, Lattanzio R, Carnevalini A, Rossi A, Coscas G. Biomicroscopy and

fluorescein angiography of pigmented iris tumors: a retrospective study on 44 cases. Int

Ophthalmol 1994;18:61–70.

14. Bashshur R, Lovett J.Assessment of telemedicine: Results of the initial experience. Aviat

Spac Environ M 1977;48:65-70.

15. Beetham WP, Aiello LM, Balodimos MC, Koncz L. Ruby-laser photocoagulation of early

diabetic neovascular retinopathy : preliminary report of long-term controlled study. Tr Am

Ophthalmol Soc 1969;67:39-67.

16. Behrendt T, Duane TD. Investigation of fundus oculi with spectral reflectance photography.

Arch Ophthalmol 1966;75:375-9.

17. Behrendt T, Wilson LA. Spectral reflectance photography of the retina. Am J Ophthalmol

1965;59:1079-88.

18. Berg TJTP, Spekreijse H. Near infrared light absorption in the human eye media. Vision Res

1997;37:249-53.

19. Bergman L, Seregard S, Nilsson B, Ringborg U, Lundell G, Ragnarsson-Olding B. Incidence

of uveal melanoma in Sweden from 1960 to 1998. Invest Ophthalmol Vis Sci 2002;43:2579-

83.

20. Bischoff PM, Flower RW. Ten years experience with choroidal angiography using

indocyanine green dye: a new routine examination or an epilogue? Doc Ophthalmol

1985;60:235-91.

21. Bjørvig S, Johansen MA, Fossen K. An economic analysis of screening for diabetic

retinopathy. J Telemed Telecare 2002;8:32-5.

22. Bloome MA. Fluorescein angiography: risks. Vision Res 1980;20:1083-97.

23. Bowe R, Char DH. Nondiagnosed uveal melanomas. Ophthalmology 2004;111: 554-7.

24. British Diabetic Association. Retinal photography screening for diabetic eye disease. A BDA

report 1997.

25. Bruun-Jensen J. Fluorescein angiography of the anterior segment. Am J Ophthalmol

1969;67:842-5.

26. Butler P, Char DH, Zarbin M, Kroll S. Natural history of indeterminate pigmented choroidal

tumors. Ophthalmology 1994;101: 710-6.

References

- 105 -

27. Bynke HG, Krakau CET. An improved stereophotographic method for clinical measurements

of optic disc protrusion. Acta Ophthalmol Scand 1960;38: 115-28.

28. Campbell FW, Boyd TAS. Use of sodium fluorescein in assessing the rate of healing in

corneal ulcers. Br J Ophthalmol 1950;34:545-9.

29. Char DH, Stone RD, Irvine AR, Crawford JB, Hilton GH, Lonn LI, Schwartz A. Diagnostic

modalities in choroidal melanoma. Am J Ophthalmol 1980;89:223-30.

30. Charman WN. Imaging in the 21st century. Ophthal Physiol Opt 1998;18:210-23.

31. Cheng H, Bron AJ, Easty D. A study of iris masses by fluorescein angiography. Trans

Ophthalmol Soc UK 1971;91:199–205.

32. Cherrick GR, Stein SW, Leevy CM, Davidson CS. Indocyanine green: observations on its

physical proportions, plasma decay, and hepatic extraction. J Clin Invest 1960;39:592-600.

33. Chopdar A. Fundus fluorescein angiography. Oxford:Butterworth-Heinemann, 1996:1-148.

34. Chopdar A, Turk AM, Hill DW. Fluorescent infrared angiography of the fundus oculi using

indocyanine green dye. Trans Ophthalmol Soc UK 1978;98:142-6.

35. Cockburn DM. Colour infrared photography of the ocular fundus. Austr J Optometry

1976;59:299-304.

36. Collaborative Ocular Melanoma Study Group. Accuracy of diagnosis of choroidal

melanomas in the Collaborative Ocular Melanoma Study. COMS Report No. 1. Arch

Ophthalmol 1990;108:1268-73.

37. Collaborative Ocular Melanoma Study Group. Factors predictive of growth and treatment of

small choroidal melanoma. COMS Report No. 5. Arch Ophthalmol 1997;115:1537-44.

38. Craandijk A, van Beek CA. Indocyanine green fluorescence angiography of the choroid. Br J

Ophthalmol 1976;60:377-85.

39. Crowther JB, Poropatich R. Telemedicine in the US army: case reports from Somalia and

Croatia. Telemed J 1995;1:73-80.

40. Cubberly MG. Infrared photography as a diagnostic tool in ophthalmology. J Biol Photogr

Assoc 1976; 44: 80-5.

41. Dart JK, Marsh RJ, Garner A, Cooling RJ. Fluorescein angiography of anterior uveal

melanocytic tumors. Br J Ophthalmol 1988;72:326–37.

References

- 106 -

42. David NJ. Infrared absorption fundus angiography. In: Proc Int Symp Fluorescein

Angiography Albi, 1969. Basel: Karger, 1971:189-92.

43. Davis MD, Fisher MR, Gangnon RE, Barton F, Aiello LM, Chew EY, Ferris FL 3rd,

Knatterud GL for the Early Treatment Retinopathy Study Research Group: Risk Factors for

high-risk proliferative diabetic retinopathy and severe visual loss. ETDRS Report Number

18. Invest Ophthalmol Vis Sci 1998;39:233-52.

44. Degenhardt AH. Light coagulation of the eye. Br J Physiol Opt 1976; 31: 11-8.

45. Dekking HM. Infrarot-Photographie des Auges. Graefe Arch Clin Exp Ophthalmol 1933;

130: 373-4.

46. Delori FC, Ben-Sira I. Excitation and emission spectra of fluorescein dye in the human ocular

fundus. Invest Ophthalmol Vis Sci 1975;14:487-92.

47. Delori FC, Gragoudas ES, Francisco R, Pruett RC. Monochromatic ophthalmoscopy and

fundus photography. The normal fundus. Arch Ophthalmol 1977;95:861-8.

48. Demeler U. Fluorescence angiographical studies in the diagnosis and follow-up of tumors of

the iris and ciliary body. Adv Ophthalmol 1981;42:1–17.

49. Dennett WS. The electric light ophthalmoscope. Trans Am Ophthalmol Soc 1885;4:156-7.

50. The Diabetic Retinopathy Study Research group. A modification of the Airlie House

classification of diabetic retinopathy. DRS report #7. Invest Ophthalmol Vis Sci 1981;21:210-

26.

51. Diener-West M, Hawkins BS, Markowitz JA, Schachat AP. A review of mortality from

choroidal melanoma. II. A meta-analysis of 5-year mortality rates following enucleation,

1966 through 1988. Arch Ophthalmol 1992;110:245-50.

52. Dieterich CE, Franz HE. Beitrag Zur Oberflächenstruktur des Pigmentepithels der

menschlichen Iris. Z Anat Entwickl-Gesch 1973;141:331-8.

53. Donaldson DD. Stereophotographic systems. Int Ophthalmol Clin 1976;16:109-31.

54. Donaldson DD. Transillumination of the iris. Trans Am Ophthalmol Soc 1974;72:89-106.

55. Douvas N, Allen L. Anterior segment photography with the Nordenson retinal camera. Am J

Ophthalmol 1950; 33: 291-2.

56. Drüner L. Ueber Mikrostereoskopie und eine neue vergrössernde Stereoskopcamera. Zeitsch

Wiss Mikrosk 1900;17:281-93.

References

- 107 -

57. Drüner L. Was ist wichtiger bei der stereoskopischen Aufnahme des vorderen

Augenabschnittes, filtriertes Licht oder Vergrösserung? Klin Monatsbl Augenheilk

1927;78:1-5.

58. Duke-Elder S, Wybar KL. The anatomy of the visual system. In:Duke-Elder S, ed. System of

Ophthalmology.Vol II. London, Kimpton. 1961:167-85.

59. Dunphy EB, Dryja TP, Albert DM, Smith TR. Melanocytic tumor of the anterior uvea. Am J

Ophthalmol 1978;86:680.

60. Early Treatment Diabetic Retinopathy Study Research Group. Early Treatment Diabetic

Retinopathy Study design and baseline patient characteristics: ETDRS report number 7.

Ophthalmology 1991;98:741-56.

61. Early Treatment Diabetic Retinopathy Study Research Group. Grading diabetic retinopathy

from stereoscopic color fundus photographs – An extension of the modified Airlie House

Classification. ETDRS report number 10. Ophthalmology 1991;98:786-806.

62. Eastman Kodak Company. Kodak filters for scientific and technical uses. Kodak publication

No. B-3, pp. 1-91. Rochester, N.Y. 1978.

63. Easty DL, Bron AJ. Fluorescein angiography of the anterior segment. Its value in corneal

disease. Br J Ophthalmol 1971;55:671-82.

64. Egan KM, Seddon JM, Glynn RJ, Gragoudas ES, Albert DM. Epidemiologic aspects of uveal

melanoma. Surv Ophthalmol 1988;32:239-51.

65. Ehrlich P. Ueber provocirte Fluorescenzerscheinungen am Auge. Deutsch Med Wochenschr

1882;8:21-35.

66. Elner VM, Flint A, Vine AK. Histopathology of documented growth in small melanocytic

choroidal tumors. Arch Ophthalmol 2004;122:1876-8.

67. Eskelin S. The development and early diagnosis of primary and disseminated uveal

melanoma. Helsinki: Yliopistopaino, 2003: 1-85.

68. Eskelin S, Kivelä T. Mode of presentation and time to treatment of uveal melanoma in

Finland. Br J Ophthalmol 2002;86:333-8.

69. The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Report of

the Expert Committee on the diagnosis and classification of diabetes mellitus. Diabetes Care

1998;21 (Suppl. 1):5-19.

References

- 108 -

70. Farman AG. Fundamentals of image acquisition and processing in the digital era. Orthod

Craniofac Res 2003;6 Suppl 1:17-22.

71. Feldman JB. A review of infra-red photography with reference to its value in ophthalmology.

Arch Ophthalmol 1936;15:435-42.

72. Fincham EF. Photographic recording of opacities of the ocular media. Br J Ophthalmol

1955;39:85-9.

73. Fine I, Loewinger E, Weinreb A, Weinberger D. Optical properties of the sclera. Phys Med

Biol 1985;30:565-71.

74. Fleiss JL, Levin B, Paik MC. Statistical Methods for Rates and Proportions. 3rd ed. New

York: John Wiley, 2003.

75. Flower RW. Extraction of choriocapillaris hemodynamic data from ICG angiograms. Invest

Ophthalmol Vis Sci 1993;34:2720-9.

76. Flower RW, Hochheimer BF. A clinical technique and apparatus for simultaneous

angiography of the separate retinal and choroidal circulations. Invest Ophthalmol Vis Sci

1973;12:248-61.

77. Flower RW, Hochheimer BF. Clinical infrared absorption angiography of the choroid. Am J

Ophthalmol 1972; 73:458-9.

78. Flower RW, Hochheimer BF. Indocyanine green dye fluorescence and infrared absorption

choroidal angiography performed simultaneously with fluorescein angiography. Johns

Hopkins Med J 1976;138;33-42.

79. Forsius H, Saari M, Nieminen H Comparative use of fluorescein angiography and redlight

and red-free light photography for examination of vascular changes of the choroid, retina and

optic nerve head. In: Francois J (ed). 5th Congress of the European Society of

Ophthalmology, Hamburg, 5-9 April 1976. Stuttgart: Ferdinand Enke Verlag, 1978:59.

80. Forster DJ, Cano MR, Green RL, Rao NA. Echographic features of the Vogt-Koyanagi-

Harada syndrome. Arch Ophthalmol 1990;108:1421-6.

81. Forster DJ, Rao NA, Hill RA, Nguyen QH, Baerveldt G. Incidence and management of

glaucoma in Vogt-Koyanagi-Harada syndrome. Ophthalmology 1993;100:613-8.

82. Freund KB, Yannuzzi LA, Sorenson JA. Age-related macular degeneration and choroidal

neovascularization. Am J Ophthalmol 1993;115:786-791.

References

- 109 -

83. Fuchs M. Provider attitudes toward STARPAHC: a telemedicine project on the Papago

reservation. Med Care 1979;17:59-68.

84. Fujimoto JG, Hee MR, Huang D, Schuman JS, Puliafito CA, Swanson E. Principles of

optical coherence tomography. In Schuman JS, Puliafito CA, Fujimoto JG (eds): Optical

coherence tomography of ocular diseases, 2nd ed. Thorofare: NJ Slack Inc, 2004: 3-19.

85. Gass JD. Observation of suspected choroidal and ciliary body melanomas for evidence of

growth prior to enucleation. Ophthalmology 1980;87:523-8.

86. Gass JDM. Pathogenesis of disciform detachment of the neuroepithelium. Am J Ophthalmol

1967;63(No 3, Part II):573-711.

87. Gass JDM. Problems in the differential diagnosis of choroidal nevi and malignant

melanomas. Am J Ophthalmol 1977;83:299-323.

88. Gass JDM. Stereoscopic atlas of macular diseases. St. Louis, Mosby 1970.

89. Geisse LJ, Robertson DM. Iris melanomas. Am J Ophthalmol 1985; 99:638–48.

90. Gocho K, Kondo I, Yamaki K. Identification of autoreactive T-cells in Vogt-Koyanagi-

Harada disease. Invest Ophthalmol Vis Sci 2001;42:2004-9.

91. Goldmann H. Spaltlampenphotographie und -photometrie. Ophthalmologica 1940;98:257-70.

92. Goldmann H. Zur Technik der Spaltlampenmikroskopie. Ophthalmologica 1938;96:90-7.

93. Gullstrand A. Die reflexlose Ophthalmoskopie. Arch Augenheilk 1911;68:101-44.

94. Gullstrand A. Demonstration der Nernst-spaltlampe. Ber Dtsch Ophthalmol Ges

1911;37:374-6.

95. Hansell P, Beeson EJG. Retinal photography in colour. Br J Ophthalmol 1953;37:65-9.

96. Harada E. Clinical observations of nonsuppurative choroiditis. Acta Soc Ophthalmol Jpn

1926;30:356.

97. Harbour JW, Augsburger J, Eagle RC. Initial management and follow-up of melanocytic iris

tumors. Ophthalmology 1995;92:1987–93.

98. Hecht J. Understanding fiber optics, 4th ed. Upper Saddle River, NJ: Prentice-Hall, 2002.

99. Hee MR, Izatt JA, Swanson EA, Huang D, Schuman JS, Lin CP, Puliafito CA, Fujimoto JG.

Optical coherence tomography of the human retina. Arch Ophthalmol 1995;113:325-32.

References

- 110 -

100. Hellstedt T, Vesti E, Immonen I. Identification of individual microaneurysms: A comparison

between fluorescein angiograms and red-free and colour photographs. Graefe Arch Clin Exp

Ophthalmol 1996 Suppl 1;234:13-7.

101. von Helmholtz H. Beschreibung eines Augenspiegels zur Untersuchung der Netzhaut in

lebenden Auge. Berlin: Forstner, 1851.

102. Helve J, Nieminen H. Simultaneous bilateral fluorescein angiography of the anterior eye.

Acta Ophthalmol Scand 1974;Suppl 123:134-5.

103. Henkind P. Book review on “Interpretation of fundus fluorescein angiography”. Am J

Ophthalmol 1978;85:882-3.

104. Henricsson M, Karlsson C, Ekholm L, Kaikkonen P, Sellman A, Steffert E, Tyrberg M.

Colour slides or digital photography in diabetes screening – a comparison. Acta Ophthalmol

Scand 2000;78:164-8.

105. Hertel E. Über Flüssigkeitsbewegungen im Auge. Ber Deutsch Ophthalmol Ges

1928;47:152-6.

106. Hoerauf H, Scholz C, Koch P, Engelhardt R, Lagua H, Birngruber R, Transcleral optical

coherence tomography. A new imaging method for the anterior segment of the eye. Arch

Ophthalmol 2002;120:816-9.

107. Hoyt WF, Newman NM. The earlierst observable sign in glaucoma? Lancet 1972;1:692-3.

108. Huang D, Li Y, Radhakrishnan S, Chalita MR. Corneal and anterior segment optical

coherence tomography. In Schuman JS, Puliafito CA, Fujimoto JG (eds): Optical coherence

tomography. Thorofare, NJ: Slack Inc, 2004: 663-73.

109. Huang D, Swanson EA, Lin CP. Optical coherence tomography. Science 1991;254:1178-81.

110. Huerkamp B. Diagnostische Möglichkeiter mit der Infrarotphotographie bei

Hornhauttrübungen. Klin Monatsbl Augenheilk 1956;128:667-72.

111. Hyvärinen L. Studies on vascular structures and circulation in the ocular fundus of normal

and hypercholesteremic rabbits. MD Thesis, University of Helsinki 1967.

112. Hyvärinen L, Flower RW. Indocyanine green fluorescence angiography. Acta Ophthalmol

Scand 1980;58:528-38.

113. Hyvärinen L, Nieminen H. Fluorescence cinematography of the ocular fundus. Acta


References

- 111 -

114. Hyvärinen L, Maumenee AE, George T, Weinstein GW. Fluorescein angiography of the

choriocapillaris. Am J Ophthalmol 1969;67:653-66.

115. Inomata H. Vogt-Koyanagi-Harada disease. In: Yanoff M, Duker JS, eds. Ophthalmology, 2nd

ed. St Louis: Mosby, 2004: 1196-9.

116. Inomata H, Rao NA. Depigmented atrophic lesions in sunset glow fundi of Vogt-Koyanagi-

Harada disease. Am J Ophthalmol 2001;131:607-14.

117. Izatt JA, Hee MR, Swanson EA, Lin CP, Huang D, Schuman JS, Puliafito CA, Fujimoto JG.

Micrometer-scale resolution imaging of the anterior eye in vivo with optical coherence

tomography. Arch Ophthalmol 1994;112:1584-9.

118. Jaeger E. Beiträgen zur Pathologie des Auges, Vol 1. Wien, 1856: 33.

119. Javitt JC, Aiello L. Cost effectiveness of detecting and treating diabetic retinopathy. Ann

Intern Med 1996;124:164-9.

120. Jensen VA, Lundbaek K. Fluorescence angiography of the iris in recent and long-term

diabetes. Congr XVIII ophthal septentr. Copenhagen 1967. Acta Ophthalmol Scand

1968;46:584-5.

121. Karma A. Ophthalmic changes in sarcoidosis. Acta Ophthalmol Scand 1979;Suppl 141:1-94.

122. Karma A, Käär M-L, Laatikainen L, Oksa H, Puomio M, Summanen P, von Wendt G, Winell

K. Suomen Diabetesliitto RY:n Retinopatiatyöryhmä. Diabeettinen retinopatia: seuranta hoito

ja näkövammautuneen diabeetikon kuntoutus. Suom Lääkäril 1992;24:2199-211.

123. Kawano Y-J, Tawara A, Nishioka Y, Suyama Y, Sakamoto H, Inomata H. Ultrasound

biomicroscopic analysis of transient shallow anterior chamber in Vogt-Koyanagi-Harada

syndrome. Am J Ophthalmol 1996;121:720-3.

124. Kersten RC, Tse DT, Anderson R. Iris melanoma. Nevus or malignancy? Surv Ophthalmol

1985; 29: 423-33.

125. Kikai K. Über die Vitalfärbung des hinteren Bulbusabschnittes. Arch Augenheilk

1930;103:541-53.

126. Kimura R, Kasei M, Shoji K, Kanno C. Swollen ciliary processes as an initial symptom in

Vogt-Koyanagi-Harada syndrome. Am J Ophthalmol 1983; 95: 402-3.

127. Kivelä T, Puska P, Raitta C, Immonen I, Tarkkanen A. Clinically successful contact

transcleral krypton laser cyclophotocoagulation. Arch Ophthalmol 1995;113:1447-53.

References

- 112 -

128. Kliman GH, Augsburger JJ, Shields JA. Association between iris color and iris melanocytic

lesions. Am J Ophthalmol 1985;100:547-8.

129. Kodolitsch FX. Infrarot-Farbphotographie im Bereigh des Auges. Klin Monatsbl Augenheilk

1969;155:284-7.

130. Kogure K, David NJ, Yamanouchi V, Choromokos E. Infrared absorption angiography of the

fundus circulation. Arch Ophthalmol 1970;83:209-14.

131. Koop N, Brinkmann R, Lankenau E, Flache S, Engelhardt R, Birngruber R. Optical

coherence tomography of the cornea and the anterior eye segment. Ophthalmologe

1997;94:481-6.

132. Kottow M. Fluorescein angiographic behaviour of iris masses. Ophthalmologica

1977;174:217–23.

133. Kottow MH. Anterior segment fluorescein angiography. Baltimore: Williams & Wilkins,

1978: 89-98.

134. Koyanagi Y. Dysakusis, Alopecia und Poliosis bei schwerer Uveitis nicht traumatischen

Ursprungs, Klin Monatsbl Augenheilkd 1929;82:194-211.

135. Krakau CET. Papillary protrusion measurements by means of stereophotographs of the

fundus. Acta Ophthalmol Scand 1956;34:140-5.

136. Kugelberg I. Der Augenhintergrund im Infrarotem Licht. Acta Ophthalmol Scand 1934;

12:179.

137. Laatikainen L. The fluorescein angiography revolution: a breakthrough with sustained

impact. Acta Ophthalmol Scand 2004;82:381-92.

138. Laatikainen L, Kohner EM. Fluorescein angiography and its prognostic significance in

central retinal vein occlusion. Br J Ophthalmol 1976;60:411-18.

139. Lentz G. Eine stereokamera zur photographie des vorderen Augenabschnittes. Klin Monatsbl

Augenheilk 1924;72:33-47.

140. Liesenfeld B, Kohner E, Piehlmeier W, Kluthe S, Aldington S, Porta M, Bek T, Obermaier

M, Mayer H, Mann G, Holle R, Hepp KD. A telemedical approach to the screening of

diabetic retinopathy: digital fundus photography. Diabetes Care 2000;23:345-8.

141. Lin DY, Blumenkranz MS, Brothers RJ, Grosvenor DM. For the Digital Diabetic Screening

group. The sensitivity and specificity of single-field nonmydriatic monochromatic digital

References

- 113 -

fundus photography with remote image interpretation for diabetic screening: A comparison

with ophthalmoscopy and standardized mydriatic color photography. Am J Ophthalmol

2002;134:204-13.

142. Littmann G. Spaltbildphotographie. Zeiss Information 1965;56:43.

143. Littmann H. Die Zeiss-Funduskamera. Ber Deutsch Ophthalmol Ges 1955;59:318-21.

144. Lubkin V. Photography of the red reflex. Arch Ophthalmol 1950;43:718-9.

145. MacLean AL, Maumenee AE. Hemangioma of the choroid. Tr Am Ophthalmol SOC

1959;57:171-94.

146. MacIlwaine WA, Anderson B, Klintworth GK. Enlargement of a histologically documented

choroidal nevus. Am J Ophthalmol 1979;87:480-6.

147. Macular Photocoagulation Study Group. Laser photocoagulation of neovascular

maculopathy: five-year results from randomized clinical trials. Arch Ophthalmol

1991;109:1109-14.

148. Maguire JI, Federman JL. Intravenous fluorescein angiography. In Tasman W, Jaeger ER

(eds) Duane’s Clinical Ophthalmology Vol. 3. Chap 4: 2000;1-38.

149. Mann WA. Color photography of the fundus oculi. Arch Ophthalmol 1932;8:405-8.

150. Mann WA. Infra-red photography of the eye. Arch Ophthalmol 1935;13:985-91.

151. Manz W. Retinitis proliferans. Graefe Arch Clin Exp Ophthalmol 1876;22:229-75.

152. Mertz M, Denffer H. Photographische und fernsehtechnische Infrarot-Untersuchung der

vorderen Augenabschnitte bei getrübter Hornhaut. Klin Monatsbl Augenheilk 1974;164:649-

60.

153. Metzger E. Die Stereophotographie des Augenhintergrundes. Klin Monatsbl Augenheilk

1927;78:338-48.

154. Meyer K, Augsburger JJ. Independent diagnostic value of fluorescein angiography in the

evaluation of intraocular tumors. Graefe Arch Clin Exp Ophthalmol 1999;237: 489-94.

155. Mims JL, Shields JA. Follow-up studies of suspicious choroidal nevi. Ophthalmology

1978;85:929-43.

156. Mitsui Y, Matsubara M, Kanagawa M. Fluorescence irido-corneal photography. Br J

Ophthalmol 1969;53:505-12.

References

- 114 -

157. Mizuno K, Majima A, Ozawa K, Ito H. Red-free light fundus photography. Invest


158. Moorthy RS, Rajeev B, Smith RE, Rao NA. Incidence and management of cataracts in Vogt-

Koyanagi-Harada syndrome. Am J Opthalmol 1994;118:197-204.

159. Moorthy RS, Inomata H, Rao NA. Vogt-Koyanagi-Harada syndrome. Surv Ophthalmol

1995;39:265-92.

160. Morax V. La Photographie stereoscopique en couleurs du segment anterieur. Ann Oculist

(Paris) 1929;166:261-4.

161. Muscat S, Parks S, Kemp E, Keating D. Secondary retinal changes associated with choroidal

naevi and melanomas documented by optical coherence tomography. Br J Ophthalmol

2004;88:120-4.

162. Nakamura H, Tsukamoto H, Shibahara R, Nagai M & Mishima. Ultrasound biomicroscopy in

the management of Vogt-Koyanagi-Harada disease. Acta Ophthalmol Scand 2000;78:718-9.

163. Nettleship E. On oedema or cystic disease of the retina. Roy Ophth Lond Hosp Rep VII

1872;3:343-50.

164. Nielsen NV. The normal fundus fluorescein angiogram and the normal fundus photograph.

Acta Ophthalmol Scand 1986; Suppl 180:1.

165. Nordenson JW. Augenkamera zum stationaren Ophthalmoskop von Gullstrand. Ber Dtsch

Ophthalmol Ges 1925;45:278.

166. Novotny HR, Alvis DL. A method of photographing fluorescence in circulating blood of the

human eye. USAF Sch Aviat Med 1960;60-82:1-4. Ref. Am J Ophthalmol 1960;50:176.

167. Novotny HR, Alvis DL. A method of photographing fluorescence in circulating blood in the

human retina. Circulation 1961;24:82-6.

168. Nussenblatt RB. Vogt-Koyanagi-Harada syndrome. In: Nussenblatt RB, Whitcup SM, eds.

Uveitis. Fundamentals and clinical practice, 3rd ed. Philadelphia: Mosby, 2004: 324-38.

169. Prasad S, Jones K, Phillips RP. Creating a district diabetic register: input from primary and

secondary care is necessary for success. Br J Gen Pract 2000;50:826.

170. Pulido JS. The requisites in ophthalmology. Retina, choroid and vitreous. St Louis: Mosby,

2002: 1-230.

References

- 115 -

171. Päivönsalo-Hietanen T, Tuominen J, Vaahtoranta-Lehtonen H, Saari KM. Incidence and

prevalence of different uveitis entities in Finland. Acta Ophthalmol Scand 1997;75:76-81.

172. Radhakrishnan S, Rollins AM, Roth JE, Yazdanfar S, Westphal V, Bardenstein DS, Izatt JA.

Real-time optical coherence tomography of the anterior segment at 1310 nm. Arch

Ophthalmol 2001;119:1179-85.

173. Raitta C. Fluorescein Angiographie nach Zentralvenenverschluss. Acta Ophthalmol Scand

1968;46:207-10.

174. Raitta C, Vannas S. Fluoresceinangiographie der Irisgefässe nach Zentralvenenverschluss.

Graefe Arch Clin Exp Ophthalmol 1969;177:33-8.

175. Raivio I. Uveal melanoma in Finland. An epidemiological, clinical, histological and

prognostic study. Acta Ophthalmol Scand 1977; Suppl 133:1-64.

176. Raivio VE, Immonen IJR, Puska PM. Transcleral red-laser cyclophotocoagulation combined

with limited anterior retinal cryocoagulation in neovascular glaucoma. Acta Ophthalmol

Scand 2007;85:60-6.

177. Read RW, Holland GN, Rao NA, Tabbara KF, Ohno S, Arellanes-Garcia L, Pivetti-Pezzi P,

Tessler HH, Usui M. Revised diagnostic criteria for Vogt-Koyanagi-Harada disease: report of

an international committee on nomenclature. Am J Ophthalmol 2001;131:647-52.

178. Read RW, Rechodouni A, Butani N, Johnston R, LaBree LD, Smith RE, Rao NA.

Complications and prognostic factors in Vogt-Koyanagi-Harada disease. Am J Ophthalmol

2001;131:599-606.

179. Redway LD. Color photography for improved teaching of ophthalmology. Am J Ophthalmol

1930;13:891-3.

180. Regillo CD. The present role of indocyanine green angiography in ophthalmology. Curr Opin

Ophthalmol 1999;10:189-96.

181. Repo LP, Suhonen MT, Teräsvirta ME, Koivisto KJ. Color doppler imaging of the

ophthalmic artery blood flow spectra of patients who have had a transient ischemic attack.

Correlations with generalized iris transluminance and pseudoexfoliation syndrome.

Ophthalmology 1995;102:1199-1205.

References

- 116 -

182. Repo LP, Teräsvirta ME, Koivisto KJ. Generalized transluminance of the iris and the

frequency of the pseudoexfoliation syndrome in the eyes of transient ischemic attack patients.

Ophthalmology 1993;100:352-5.

183. Repo LP, Teräsvirta ME, Tuovinen E. Generalized peripheral iris transluminance in the

pseudoexfoliation syndrome. Ophthalmology 1990;97:1027-9.

184. Retinopathy Working Party. A protocol for screening for diabetic retinopathy in Europe.

Diabetic Medicine 1991;8:263-7.

185. Roberts DK. Digital infrared photography assists in the detection of iris and ciliary body

cysts. J Glaucoma 2003; 12:288-91.

186. Romanchuk KG. Fluorescein. Physicochemical factors affecting its fluorescence. Surv

Ophthalmol 1982;26:269-83.

187. Rootman J, Gallagher RP. Color as a risk factor in iris melanoma. Am J Ophthalmol

1984;98:558.

188. Rosen ES, Fluorescence photography of the eye. London: Butterworths, 1969.

189. Rubsamen PE, Gass JDM. Vogt-Koyanagi-Harada syndrome. Arch Ophthalmol

1991;109:682-7.

190. Rutzen AR, Ortega-Larrocea G, Frambach DA, Rao NA. Macular edema in chronic Vogt-

Koyanagi-Harada syndrome. Retina 1995;15:475-9.

191. Ryan SJ, ed. Retina, 4th ed. Philadelphia: Elsevier, 2006: 1-2654.

192. Saari JM, Kerola MT, Broas M, Saari KM. Hand-held digital imaging video camera for eye

examination and follow-up. J Telemed Telecare 2002;8:237-240.

193. Saari KM. Anterior segment fluorescein angiography in inflammatory diseases of the cornea.

Acta Ophthalmol Scand 1979;57:781-93.

194. Saari M. Disciform detachment of the macula. II. Fluorescein and indocyanine green

fluorescence angiographic findings in juvenile hemorrhagic macular choroidopathy. Acta


195. Saari M. Vogt-Koyanagi-Harada syndrome. Report of a case with many recurrences.

Ophthalmologica 1974;169:326-32.

196. Saari M, Nieminen H. Fluorescein angiography and infrared transillumination stereo

technique for studying the ciliary body and iris. In: J. François, ed. 5th Congress of the

References

- 117 -

European Society of Ophthalmology, Hamburg, 5-9 April 1976. Stuttgart: Ferdinand Enke

Verlag, 1978:46-8.

197. Saari M, Vuorre I, Kaila J, Lahti R. Family studies of ocular manifestations in arthritis. Can J

Ophthalmol 1978a;13:144-51.

198. Saari M, Vuorre I, Nieminen H. Fuchs’s heterochromic cyclitis: A simultaneous bilateral

fluorescein angiographic study of the iris. Br J Ophthalmol 1978b;62:715-21.

199. Saari M, Vuorre I, Nieminen H. Infrared transillumination stereophotography of normal iris.

Can J Ophthalmol 1977;12:308-11.

200. Saari M, Vuorre I, Nieminen H. Infrared transillumination stereophotography of the iris in

Fuchs’s heterochromic iridocyclitis. Br J Opthalmol 1978c;62:110-5.

201. Saari M, Vuorre I, Nieminen H, Räisänen S. Acquired toxoplasmic chorioretinitis. Arch

Ophthalmol 1976;94:1485-8.

202. Saari M, Vuorre I, Räisänen S, Nieminen H. Vascular changes in toxoplasmic chorioretinitis.

In: J. Francois, ed. 5th Congress of the European Society of Ophthalmology, Hamburg, 5-9

April 1976. Stuttgart: Ferdinand-Enke Verlag, 1978: 662-4.

203. Saine PJ, Tyler ME. Ophthalmic photography: retinal photography, angiography, and

electronic imaging, 2nd ed. Boston: Butterworth-Heinemann, 2002: 1-321.

204. Samei E, Dobbins JT, Lo JY, Tornai MP. A framework for optimising the radiographic

technique in digital X-ray imaging. Radiat Prot Dosimetry 2005;114:220-9.

205. Sardar DK, Swanland G-Y, Yow RM, Thomas RJ, Tsin ATC. Optical properties of ocular

tissues in the near infrared region. Lasers Med Sci 2007;22(1):46-52.

206. Schatz H, Burton TC, Yannuzzi LA, Rabb MF. Interpretations of fundus fluorescein

angiography. St. Louis: Mosby, 1978: 1-761.

207. Schuman JS, Puliafito CA, Fujimoto JG. Optical coherence tomography of ocular diseases,

2nd ed. Thorofare, NJ: Slack Inc, 2004: 1-714.

208. De Schweinitz GE, Shumway EA. Concerning melanoma of the choroid, with report of one

case of this character and another exhibiting a pigmented sarcoma of the choroid early in its

development. Trans Am Ophthalmol Soc 1905;10:439-51.

References

- 118 -

209. Seddon JM, Sahagian CR, Glynn RJ, Sperduto RD, Gragoudas ES, for the Eye disorders

Case-Control Study Group. Evaluation of an iris color classification system. Invest

Ophthalmol Vis Sci 1990;31:1592-8.

210. Sharp PF, Olson J, Strachan F, Hipwell J, Ludbrook A, O’Donnel M, Wallace S, Goatman K,

Grant A, Waugh N, McHardy K, Forrester JV. The value of digital imaging in diabetic

retinopathy. Health Technology Assessment 2003;7:pp. 14.

211. Shiba T, Maruo K, Akahoshi T. Development of a multi-field fundus photographing system

using a non-mydriatic camera for diabetic retinopathy. Diabetes Res Clin Pract 1999;45:1-8.

212. Shields CL, Demirci H, Materin MA, Marr BP, Mashayekhi A, Shields JA. Clinical factors in

the identification of small choroidal melanoma. Can J Ophthalmol 2004;39:351-7.

213. Shields CL, Shields JA. Clinical features of small choroidal melanoma. Curr Opin

Ophthalmol 2002;13:135-41.

214. Shields CL, Shields JA, Kiratli H, de Potter P, Cater JR, Mclean IW. Risk factors for growth

and metastasis of small choroidal melanocytic lesions. Ophthalmology 1995;102:1351-61.

215. Shields JA. Diagnosis and management of intraocular tumors. St. Louis: C. V. Mosby, 1983:

83-121.

216. Shields JA, Augsburger JJ, Brown GC, Stephens RF. The differential diagnosis of posterior

uveal melanoma. Ophthalmology 1980;87:518-22.

217. Shikano S, Shimizu K. Atlas of fluorescence fundus angiography. Philadelphia: WB

Saunders, 1968: 1-201.

218. Shimmura S, Shinozaki N, Fukagawa K, Shimazaki J, Tsubota K. Real-time telemedicine in

the clinical assessment of the ocular surface. Am J Ophthalmol 1998;125:388-90.

219. Shindo Y, Inoko H, Yamamoto T, Ohno S. HLA-DRB1 typing of Vogt-Koyanagi-Harada’s

disease by PCR-RFLP and the strong association with DRBI*0405 and DRB1* 0410. Br J


220. Singh AD, Kivelä T. The collaborative ocular melanoma study. Ophthalmol Clin N Am

2005;18:129-42.

221. Singh AD, Mokashi AA, Bena JF, Jacques R, Rundle PA, Rennie IG. Small choroidal

melanocytic lesions. Ophthalmology 2006;113:1032-9.

222. Statistics Finland. Statistical Yearbook of Finland 1989-2004. Serial VI, volumes 84-99.

References

- 119 -

223. Stokes ME, Davis CS, Koch GG. Categorical data analysis using the SAS system. Cary, NC:

SAS Institute Inc, 1995.

224. Summanen P, Kallioniemi V, Komulainen J, Lamminen H, Levänen H, Orhanen E, Tulokas

S, Virtanko T, von Wendt G. Diabeettinen retinopati. Duodecim 2007;123:421-2.

225. Sunba MSN, Rahi AHS, Morgan G. Tumors of the anterior uvea. I. Metastasizing malignant

melanoma of the iris. Arch Ophthalmol 1980;98:82-5.

226. Sysi R. A simple device supplementary to the slit lamp to photograph the parts of the eye.

Acta Ophtalmol Scand 1949;27:403-8.

227. Taylor R. Practical community screening for diabetic retinopathy using the mobile retinal

camera: report of a 12 centre study. Diabet Med 1996;13:946-52.

228. Territo C, Augsburger JJ, Schroeder RP, Shields JA. Enlargement of melanocytic iris lesions.

Ophthalmic Surg 1987;18:644-9.

229. Territo C, Shields CL, Shields CL, Shields JA, Augsburger JJ, Schroeder RP. Natural course

of melanocytic tumors of the iris. Ophthalmology 1998;95:1251–5.

230. Thiel R. Photographierte Spaltlampenbilder. Ber Dtsch Ophthalmol Ges 1930;48:355-7.

231. Thorner W. Die stereoskopische Photographie des Augenhintergrundes. Klin Monatsbl

Augenheilk 1909;47:481-90.

232. Tyler ME, Saine PJ, Bennett TJ. Practical retinal photography and digital imaging

techniques. Philadelphia: Butterworth-Heineman, 2003: 1-222.

233. Vannas A. Fluorescein angiography of the vessels of the iris in pseudoexfoliation of the lens

capsule, capsular glaucoma and some other forms of glaucoma. Acta Ophthalmol Scand

1969;Suppl 105:1-75.

234. Vannas A, Raitta C, Lemberg S. Photography of the nerve fiber layer in retinal disturbances.

Acta Ophthalmol Scand 1977; 55:79-87.

235. Vannas S, Raitta C. Retinal circulation in high blood pressure. Int Ophthalmol Clin

1968;8:291-335.

236. Vogel A, Dlugos C, Nuffer R, Birngruber R. Optical properties of Human Sclera and Their

Consequences for Transscleral Laser Applications. Laser Surg Med 1991;11:331-340.

237. Vogt A. Frühzeitiges Ergrauen der Zilien und Bemerkungen üben den sogenannten

plözlichen Eintritt dieser Verärderung. Klin Monatsbl Augenheilkd 1906;44:228-42.

References

- 120 -

238. Vogt A. Herstellung eines goldblauer Lichtfiltrats, in welchem die Macula Centrales in Vivo

in gelber Färbung erscheint, die Nervenvasern der Netzhaut und andere feine Einzelheiten

derselben sichtbar warden, und der Grad der gelbfärbung der Linse ophthalmoskopisch

nachweisbar ist. Graefe Arch Clin Exp Ophthalmol 1913;84:293.

239. Voutilainen-Kaunisto R, Niskanen L, Uusitupa M, Teräsvirta M. Iris transluminance in type

2 diabetes. Acta Ophthalmol Scand 2002;80:64-8.

240. Webb RH, Hughes GW, Delori FG. Confocal scanning laser ophthalmoscope. Appl Opt

1987;26:1492-9.

241. Webb RH, Hughes GW, Pomerantzeff O. Flying spot TV ophthalmoscope. Appl Opt

1980;19:1991-7.

242. von Wendt G. Screening for diabetic retinopathy. Aspects of photographic methods. MD

Thesis, Karolinska institutet, Stockholm. Universitetsservice US AB 2005: 1-54.

243. von Wendt G, Heikkilä K, Summanen P. Assessment of diabetic retinopathy using two-field

60° fundus photography. A comparison between red-free black-and-white prints and colour

transparencies. Acta Ophthalmol Scand 1999;77:638-47.

244. von Wendt, Heikkilä K, Summane n P. Detection of retinal

neovascularizations using 45° and 60° photographic fields with varying 45° fields simulated

on a 60° photograph. Acta Ophthalmol Scand 2002;80:372-8.

245. von Wendt G, Rönnholm P, Heikkilä K, Summanen P. A comparison between one- and two-

field 60° fundus photography when screening for diabetic retinopathy. Acta Ophthalmol

Scand 2000;78:14-20.

246. Wessing A, Meyer-Schwickerath G. Fluoreszenzangiographie der Retina. Stuttgart: George

Thieme Verlag, 1968: 1-212.

247. Wheeler HO, Cranston WI, Mettzer JI. Hepatic uptake and biliary excretion of indocyanine

green in the dog. Proc Soc Exp Biol Med 1958;99:11-4.

248. Wilson RS. Does the lens diameter increase or decrease during accommodation? Human

accommodation studies: infrared retro-illumination video photography and pixel unit

measurements. Trans Am Ophthalmol Soc 1997;95:261-7.

249. Wise GN, Dollery CT, Henkind P. The retinal circulation. New York: Harper & Row, 1971:

1-566.

References

- 121 -

250. World Health Organization (WHO) Mass health examinations. Publ health papers N:o 45,

Geneva: World Health Organization, 1971.

251. Yannuzzi LA, Ober MD, Slakter JS, Spaide RF, Fisher YL, Flower RW, Rosen R.

Ophthalmic fundus imaging: Today and beyond. Am J Ophthalmol 2004;137:511-24.

252. Yannuzzi LA, Slakter JS, Sorenson JA, Guyer DR, Orlock DA. Digital indocyanine green

videoangiography and choroidal neovascularization. Retina 1992;12:191-223.

253. Yogesan K, Constable IJ, Barry CJ, Eikelboom RH, McAllister IL, Tay-Kearney ML.

Telemedicine screening of diabetic retinopathy using a hand-held fundus camera. Telemed J

2000;6:219-23.

254. Yogesan K, Constable IJ, Barry CJ, Eikelboom RH, Morgan W, Tay-Kearney ML, Jitskaia L.

Evaluation of a portable fundus camera for use in the teleophthalmologic diagnosis of

glaucoma. J Glaucoma 1999;8:297-301.

255. Yogesan K, Constable IJ, Eikelboom RH, van Saarloos PP. Teleophthalmic screening using

digital imaging devices. Aust NZ J Ophthalmol 1998;26 Suppl. 1:9-11.

Documents

Digital Photography in the Diagnosis and Follow-up of ... · atrophy and serous detachment of the ciliary body with ocular hypotony together with the shallow retinal detachment of