10/29/2012
1
Clinical Applications of Corneal Optical Coherence
Frontiers in Optics/Laser Science XXVIIIRochester, NY, Oct 14-18, 2012
Corneal Optical Coherence Tomography
David Huang, MD, PhDWeeks Professor of Ophthalmic Research
Professor of Ophthalmology & Biomedical Engineering
Casey Eye Institute, Oregon Health & Science University
Portland, Oregon
Financial Interests:Dr. D. Huang has a significiant financial interest in Optovue, a company that may have a commercial interest in the results of this research and technology. This potential individual conflict of interest has been reviewed and managed by OHSU.Optovue, Inc.: stock options, patent royalty, grants, speaker honorarium & travel supportCarl Zeiss Meditec, Inc.: patent royalty
New OCT Technology Lead to An Explosion of New Products for Anterior Segment Imaging
Zeiss Cirrus Optovue RTVue
Bioptigen Envisu
Zeiss Visante Optovue iVue
Heidelberg Spectralis
Optopol Copernicus
Heidelberg SL-OCT
Tomey CASIA
David Huang, MD, PhD www.COOLLab.net
10/29/2012
2
Anterior Segment OCT Systems
Dedicated systems scan wider (12-16 mm)
Systems converted from retinal scanners have
and deeper (6-7 mm)
Zeiss Visante
Heidelberg SL-OCT
Tomey CASIA
limited scan ranges (~6mm width & 2mm depth)
Optovue RTVue
Optovue iVue
Zeiss Cirrus
H id lb S t li Heidelberg Spectralis
Bioptigen Envisu
Optopol Copernicus
David Huang, MD, PhD www.COOLLab.net
Wavelength Matters
1310 nm gives deeper penetration.
830 nm systems have higher axial resolution.
(Axial resolution: 17-18 µm)
Zeiss Visante
Heidelberg SL-OCT
Tomey CASIA
(Axial resolution: 3-5 µm)
Optovue RTVue
Optovue iVue
Zeiss Cirrus
Heidelberg Spectralis
Bioptigen Envisu Bioptigen Envisu
Optopol Copernicus
David Huang, MD, PhD www.COOLLab.net
10/29/2012
3
Better penetration at 1310 nm Greater resolution achieved w/ 830 nm systems
830 nm1310 nm
Schwalbe’s line
Scleral spur
Epithelium
Endothelium
Angle recess
David Huang, MD, PhD www.COOLLab.net
Visante TD-OCT RTVue FD-OCT
Fourier vs. Time Domain
Time domain systems are slower
Fourier domain systems are much faster
Zeiss Visante (2 kHz)
Heidelberg SL-OCT (200 Hz)
Optovue RTVue (26 kHz)
Optovue iVue (26 kHz)
Zeiss Cirrus (27 kHz)
Heidelberg Spectralis (40 kHz)
Bi ti E i (17 kH ) Bioptigen Envisu (17 kHz)
Optopol Copernicus (20 kHz)
Tomey CASIA (30 kHz)
David Huang, MD, PhD www.COOLLab.net
10/29/2012
4
Clinical Applications of Corneal OCT
Keratoconus DetectionLASIK planning FS Laser AK
Guiding PTK IOL Power Calculation Intacs
David Huang, MD, PhD www.COOLLab.net
Keratoconus Diagnosis with OCT
Bing Qin, MD, PhD
Yan Li, PhD
David Huang, MD, PhD www.COOLLab.net
10/29/2012
5
Forme fruste keratoconus is the most important risk factor for post-LASIK corneal ectasiaBinder P, Analysis of ectasia after LASIK: risk factors, J Cat Refract Surg 33:1530, 2007
T h N l S i i Ab l T t l
Topography does not screen out all eyes at risk
Randleman JB et al. Risk assessment for ectasia after corneal refractive surgery. Ophthalmol 2007; 115:37-50
Topography Normal Suspicious Abnormal Total
# Ectasia Cases
25 27 33 93
Percent 27% 29% 41% 100%
OCT could provide more accurately map corneal thickness, epithelial thickness, and corneal cuvatures
1.5mm 1.5mm
Rp Ra
Dn0 = 1
n1 = 1.376
n2 = 1.336
aa R
nnK 01
pp R
nnK 12
David Huang, MD, PhD www.COOLLab.net
RTVue (Optovue, Inc.)
10/29/2012
6
Corneal Mapping over 8 Meridians(8 x 1019 A-scans) in 0.31 Second
RTVue FD-OCT(Optovue, Inc.)
David Huang, MD, PhD www.COOLLab.net
Detecting Keratoconic Thinning with OCT Pachymetric Indices
General thinning Min Min
Focal thinning Minimum - median
Asymmetric thinning I-S IT-SN
Minimum = 404 µmY = -710 µm
Y location of the Minµ
Li Y, Meisler M, Tang M, Lu A, Thakrar V, Reiser B, Huang D. Keratoconus diagnosis with optical coherence tomography pachymetry mapping. Ophthalmology 2008;115:2159-2166.
Visante Prototype (Carl Zeiss Meditec, Inc.)
10/29/2012
7
OCT Keratoconus Pachymetric VariablesParameter Explanation Keratoconus Cutoff
(1 percentile)IT-SN Average thickness of the IT octant minus that of the SN
octant< -51 μm
I-S Average thickness of the inferior (I) octant minus that of the superior (S) octant
< -49 μm
Min Minimum corneal thickness < 455 μm
Min - Med Minimum corneal thickness-median corneal thickness <-29 μm
Y Min Y coordinate of minimum corneal thickness Y < -1.35 mm
Based on 66 normal and 52 keratoconus subjects from the
Yan Li, PhD; David Huang, MD, PhD www.COOLLab.net
Based on 66 normal and 52 keratoconus subjects from the following:
David Huang, MD, PhD, Doheny Eye Institute, Los AngelesQinmei Wang, MD, Wenzhou Medical College, WenzhouRobert Brass, MD, Brass Eye Center, New York
RTVue FD-OCT (Optovue, Inc.)
OCT pachymetry map-based keratoconus score table
0 1 2 3 OD OS
Minimum >499 499 ~ 476 475 ~ 455 <455Minimum 499 499 476 475 455 455
Minimum- Median >-21 -22 ~ -25 -26 ~ -29 <-29
I-S >-30 -30~-40 -41~-49 <-49
IT-SN >-33 -33~42 -43~-51 <-51
Y location of the min >-734 -734 ~ -1069 -1070 ~ -1353 <-1353
Sum
mat
ion
Keratoconus Risk Score
Each variable will be assigned a score of 1, 2, 3 if it exceeds 20, 5, 1 percentile thresholds.The keratoconus risk score of the eye is the summation of all scores.
Keratoconus risk score = 0~1: low risk; 2~3: moderate risk; ≥4: high risk.
Yan Li, PhD; David Huang, MD, PhD www.COOLLab.net
10/29/2012
8
Histogram of the Keratoconus Score
80
90 Diagnostic criterion:Keratoconus score ≥ 4
20
30
40
50
60
70
80
Normal
Keratokonus
Num
bero
f Eye
s
Sensitivity = 86% Specificity = 94%
AROC = 0.951
0
10
0 1 2 3 >=4
Yan Li, PhD; David Huang, MD, PhD www.COOLLab.net
Epithelial Thinning Over Cone
Epithelial Thickness Pachymetry Axial Power
58 µm47 µm
Epithelial mapping software pending FDA Approval
58 µm47 µm
Yan Li, PhD; David Huang, MD, PhD www.COOLLab.net
10/29/2012
9
Epithelial Thickness Map by FD-OCT
Magnified corneal OCT image Epithelial thickness map
Central
D=2 5 mm
Superior
Inferior
2 group of subjects: normal and keratoconus Exclusion criteria: eyes with late keratoconic
changes such as scar or hydrops
D=2,5 mm
Yan Li, PhD; David Huang, MD, PhD www.COOLLab.net
Epithelial Map Repeatability
N Central Superior Inferior
Normal 103 0.7 µm 0.8 µm 0.7 µm
Keratoconus 35 1.0 µm 1.0 µm 1.0 µm
pooled standard deviation (SD)
Li Y, Tan O, Brass R, Weiss JL, Huang D. Corneal epithelial thickness mapping by Fourier-domain optical coherence tomography. Ophthalmology; 2012 Aug 20, Epub ahead of print.
10/29/2012
10
Average epithelial thickness maps show inferotemporal thinning
Normal
thinning over apexthinning over apex
Keratoconus
Yan Li, PhD; David Huang, MD, PhD www.COOLLab.net
Epithelial Map Variables Minimum epithelial thickness in µm
Minimum – maximum (MIN-MAX) focal thinning in µm
Map standard deviation (MSD) in µm Map standard deviation (MSD) in µm
Pattern standard deviation (PSD)
Minimum MIN-MAX MSD PSD
Normal 45.9 ± 4.7 -8.9 ± 3.4 2.1 ± 0.7 0.031±0.008
40 0 6 0* 18 8 0* 4 2 0* 0 10 0 030*Keratoconus 40.0±6.0* -18.7±8.0* 4.7 ± 2.0* 0.105±0.030*
AROC 0.84 0.88 0.89 1.0
Yan Li, PhD; David Huang, MD, PhD www.COOLLab.net
* Statistically significant difference was detected between normal and keratoconic eyes.AROC = area under the receiver operating characteristic curve.
10/29/2012
11
Pattern Standard Deviation (PSD) of epithelial map was highly accurate in forme fruste keratoconus (FFK) detection
Pilot study 17 FFK eyes
17 normal eyes
AROC = 1 0 for FFK detection AROC = 1.0 for FFK detection 100% sensitivity
100% specificitivity
Yan Li, PhD; David Huang, MD, PhD www.COOLLab.net
Conclusions
OCT measures many variables that provide accurate keratoconus diagnosis Pachymetry map variables
Epithelial thickness map variables
Anterior curvature
Posterior curvature
David Huang, MD, PhD www.COOLLab.net
10/29/2012
12
OCT for the LASIK surgeon
Camila Salaroli, MD
Jose Luiz B. Ramos, MD
Yan Li, PhD
David Huang, MD, PhD www.COOLLab.net
Flap thickness varies from setting
138-165 µm
RS035KA, OD
IntraLase 120 µm setting 117-131 µm
RSAA, OS
David Huang, MD, PhD www.COOLLab.net
10/29/2012
13
OCT Flap & Bed Thickness Measurements are Precise
1 week post-iLASIK
Position -2,5mm -1,5mm -0,5mm +0,5mm +1,5mm +2,5mm
POOLED SD (μm)
2.3 1.3 3.1 3.0 2.0 2.3
Consecutive series of 17 eyes of 10 patients, Intralase 110 m settingDavid Huang, MD, PhD www.COOLLab.net
Routinely use OCT to check flap thickness one week after LASIK and calculate statistical upper bound
Post-LASIK 1 week
• Intralase 110 µm setting (my surgeries)• OCT measurements in 21 eyesOCT measurements in 21 eyes
• Mean = 124 µm • standard deviation (SD) = 7 µm
• Upper bound = mean + 2 SD = 138 µm• < 5% of flaps will exceed this thickness
David Huang, MD, PhD www.COOLLab.net
10/29/2012
14
Before enhancement, use OCT to predict residual stromal bed thickness
Residual bed thickness = 451 – 143 – 37 = 271 µm
OCT map minimumCentral flap thickness
Ablation depth
David Huang, MD, PhD www.COOLLab.net>250
Post-LASIK interface fluid & epithelial ingrowth
Epithelial ingrowth Fluid
Fibrosis
Epithelial ingrowth Fluid
Ramos JLB, Zhou S, Yo C, Tang M, Huang D, High-resolution imaging of complicated LASIK flap interface fluid syndrome. Ophthalmic Surg Lasers Imaging 2008;39:S80-2
10/29/2012
15
Planning Phototherapeutic Keratectomy
Nehal Samy, MDCatherine Cleary, MD Yan Li, PhD
David Huang, MD, PhD www.COOLLab.net
Not A Candidate for PTK
Full thickness scar -> PK needed
David Huang, MD, PhD www.COOLLab.net
10/29/2012
16
Nodular degeneration is peeled
Nodular scar above Bowman’s layerStromal scar peripheral
Peel and then polish using laser with masking agent
David Huang, MD, PhD www.COOLLab.net
Example: Granular Corneal Dystrophy
Before PTK
David Huang, MD, PhD www.COOLLab.net
UCVA: 20/80-1 BSCVA: 20/50-2MR: -2.00+2.00x100° Sim-K: 42.2/44.8@98°
10/29/2012
17
PTK SimulationAblation map
•Transepithelial PTK
6.0 mm AZ circle, 64µm
•Refractive correction
-3D+2.2Dx100°
6mmx4.5mm
Yan Li, PhD; David Huang, MD, PhD www.COOLLab.net
Simulation software is experimental and not commercially available
Comparison of Simulated and Actual PTK Outcome en face
Before PTK
Actual OutcomeSimulated OutcomePost-PTK Slitlamp photoYan Li, PhD; David Huang, MD, PhD www.COOLLab.net
10/29/2012
18
Comparison of simulated and actual PTK outcome in cross-section
Before PTK
Simulated Outcome445 µm
531 µm505 µm
544 µm619 µm593 µm
Actual Outcome
David Huang, MD, PhD www.COOLLab.net
2.5 mm2.5 mm
446 µm497 µm 519 µm
OCT ablation efficiency analysis
VISX S4 excimer laser
Laser ablation efficiency = actual ablation depth / nominal ablation setting
Central (average within 1 mm diameter): 120%
Periphery (5 mm diameter ring): 126%
David Huang, MD, PhD www.COOLLab.net
10/29/2012
19
Epithelial thickness profile predicts spherical equivalent refractive change
Refractive change = -0.29 + 0.141*(PTK depth - CET) - 0.159*(CET-PET) –refractive ablation setting
CET = central epithelial thickness
PET = peripheral epithelial thicknessDavid Huang, MD, PhD www.COOLLab.net
OCT is Useful in PTK Planning
Is PTK the best option?Deep opacity > keratoplasty is needed Deep opacity -> keratoplasty is needed
Opacity above Bowman’s layer -> peel
Plan transepithelial ablation Measure opacity depth
Measure actual laser ablation depth
Calculate ablation needed to remove opacity
Measure epithelial thickness distribution
Predict refractive outcome
David Huang, MD, PhD www.COOLLab.net
10/29/2012
20
Maolong Tang, PhD
Intraocular Lens Power Calculation
David Huang, MD, PhD www.COOLLab.net
The problem with keratometry after laser vision correction
Measurement
Pre-LASIKExtrapolation
Post-LASIK
The extrapolation no longer works!David Huang, MD, PhD www.COOLLab.net
10/29/2012
21
OCT-based IOL Power Formula
Position Vergence (D)
Back of IOL V1 = 1000 / (AL-ELP-CCT-0.15)/n2
Front of IOL V1’ = V1 – IOL power
IOL power ECP
Back of cornea V2 = 1000/ [1/V1’ + (ELP+CCT)/n2]
Front of cornea V2’ = V2 – ECP
n2 = refractive index of vitreous and aqueous (1.336)
AL = axial eye length (mm)
ELP ff ti l iti ( )
V1 V1’ V2 V2’
ELP = 0.711 * (ACD-CCT) – 0.25 * Pp + 0.623 * ALadj + pACD – 8.11ACD = Anterior chamber depth (mm)
CCT = Central corneal thickness (mm)Pp = posterior corneal power (D)
ALadj = sqrt(AL) if AL < 24.4mm= sqrt(AL+0 8*(AL 24 4)) if AL > 24 4mmELP = effective lens position (mm)
CCT = central corneal thickness (mm)
ECP, effective corneal power (D)
= 1.0208 * net corneal power – 1.6622 (for myopic laser vision correction)2
= 1.1 * net corneal power – 5.736 (for hyperopic laser vision correction)3
1. Tang M, Li Y, Huang D. An intraocular lens power calculation formula based on optical coherence tomography: a pilot study. J Refract Surg 2010; 26:430-437 2. Tang M, Wang L, Koch DD, Li Y, Huang D, Intraocular Lens Power Calculation after Previous Myopic Laser Vision Correction Based on Corneal Power Measured by
Fourier-Domain Optical Coherence Tomography, J Cataract Refract Surg. In press.3. Tang M, Wang L, Koch DD, Li Y, Huang D, Intraocular Lens Power Calculation after Myopic and Hyperopic Laser Vision Correction Using Optical Coherence Tomography.
Saudi Journal of Ophthalmology. 2012 Feb; 26: 19-24
= sqrt(AL+0.8 (AL-24.4)), if AL > 24.4mmpACD = ACD-constant
= [(A-constant*0.5663)-65.6+3.595]/0.9704
Input VariablesPartial coherence biometer
(e.g. IOL-Master, Carl Zeiss Meditec) or ultrasound A-scan
1) ACD – anterior chamber depth
2) AL – axial eye length
3) Net corneal power
4) Anterior corneal power
5) Posterior corneal power
6) Central corneal thicknessFourier-domain OCT
(RTVue, Optovue, Inc.)
Maolong Tang, PhD; David Huang, MD, PhD www.COOLLab.net
10/29/2012
22
Prospective clinical study on post-laser vision correction cataract surgery
Clinical SiteMyopic LVC
(eyes)Hyperopic LVC
(eyes)(eyes) (eyes)
Cullen Eye Institute 12 6
Doheny Eye Institute 10 3
Total 22 9Total 22 9
LVC = laser vision correction
Postoperative manifest refraction spherical equivalent was evaluated at 1 month
Tang M, Wang L, Koch DD, Li Y, Huang D, Intraocular Lens Power Calculation after Myopic and Hyperopic Laser Vision Correction Using Optical Coherence Tomography. Saudi Journal of Ophthalmology. 2012 Feb; 26: 19-24
Accuracy of OCT-based IOL power calculation compared with Haigis-L after myopic LVC
Keratometry Method
IOL Formula
Prediction Error (D)
Range (D)
MAE (D)
Within 1D
IOL-Master Haigis-L -0 35 ± 0 94 (-2 32 1 30) 0 73 73%O aste a g s -0.35 ± 0.94 (-2.32, 1.30) 0.73 73%
OCT OCT 0.07 ± 0.80 (-1.24, 2.15) 0.57 82%
p = 0.19. n = 22 eyes of 16 subjects.
Maolong Tang, PhDDavid Huang, MD, PhD
www.COOLLab.net
10/29/2012
23
Accuracy of OCT-based IOL power calculation compared with other methods after myopic LVC at the Doheny Eye Institute
Keratometry IOL Formula
Prediction Range MAEMethod
IOL FormulaError (D) (D) (D)
Clinical History Hoffer-Q 0.31 ± 2.52 -1.47 to 2.09 1.78*
CL Overrefraction Holladay II 0.01 ± 2.08 -3.09 to 3.19 1.46*
IOL-Master Haigis-L -0.28 ± 1.19 -2.32 to 0.98 0.88
Orbscan II Holladay II 0.17 ± 1.61 -2.72 to 1.51 1.28
OCT OCT-based 0.49 ± 0.64 -0.33 to 1.58 0.60
*p<0.05 compared to OCT. N = 10 eyes.The standard IOL formula that provided the smallest MAE was used for all keratometry methods except OCT.
OCT-based IOL calculation was significantly better than clinical history and contact lens over-refraction methods.
Maolong Tang, PhD, David Huang, MD, PhD www.COOLLab.net
Conclusion
The predictive accuracy of OCT-based IOL power calculation is equal to or better than p qcurrent standards for post-LVC eyes.
More improvement are possible Aberration measurement or ray tracing
Better prediction of IOL position
Prediction of astigmatism outcomeg
Maolong Tang, PhD, David Huang, MD, PhD www.COOLLab.net
10/29/2012
24
Grant & Material Supports
NIH R01 EY018184: Guiding The gTreatment of Anterior Eye Diseases with Optical Coherence Tomography
Unrestricted grant from Research to Prevent BlindnessPrevent Blindness
Grant & material support from Optovue, Inc.
www.COOLLab.net
David Huang, MD, PhD
Maolong Tang, PhD
Ou Tan, PhD
Yimin Wang, PhD
Yan Li, PhD
Xinbo Zhang, PhD
Jason Tokayer, MS
Janice Van Norman, COT
Yali Jia, PhD
Kathleen S. Torok, MA
10/29/2012
25
New Technologies from The Laboratory
HIGHER SPEED GREATER DEPTH RANGE
Anterior Segment Imaging and Extraction of Biometric Parameters
• James G. Fujimoto, Massachusetts Institute of Technology
Volumetric OCT data
OCT registration
3D refraction correction
3D OCT biometry
10/29/2012
26
Pachymetry Map
I. Grulkowski et al., Volumetric anterior segment biometry using OCT registration and refraction correction techniques, manuscript in preparation
Topography
I. Grulkowski et al., Volumetric anterior segment biometry using OCT registration and refraction correction techniques, manuscript in preparation
10/29/2012
27
360° Anterior Chamber Angle Survey
I. Grulkowski et al., Volumetric anterior segment biometry using OCT registration and refraction correction techniques, manuscript in preparation
Anterior Segment Imaging Using VCSEL
• VCSEL technology
• Long coherence length
1 mm
I. Grulkowski et al., Retinal, anterior segment and full eye imaging using ultrahigh speed swept source OCT with vertical-cavity surface emitting lasers, Biomed. Opt. Express 3, 2733 (2012).
10/29/2012
28
Full Eye Imaging for Non-Contact Ocular Biometry
Cornea Intraocular DistancesCentral Corneal
0 527±0 003mm
50ms
Crystalline Lens
Thickness0.527±0.003mm
Aqueous Depth 3.305±0.030mm
Anterior Chamber Depth
3.831±0.029mm
Lens Thickness 3.880±0.030mm
Vitreous Depth 18.674±0.018mm
Axial Eye
5 mm
Retina
Axial Eye Length
26.384±0.016mm
IOL Master – 20 umImmersion Ultrasound – 100 um
Speckle free imaging – cornea in-vivo
a.
a. Epithelium & Bowman’s Layer
b. b. Endothelium
c. c.Epithelium & Bowman’s Layer
d.d.
Endothelium
Courtesy of Maciej Wojtkowski, Nicolaus Copernicus University, Torun, Poland
M. Szkulmowski,et al. "Efficient reduction of speckle noise in Optical Coherence Tomography," Opt. Express 20, 1337, 2012
10/29/2012
29
Speckle free imaging – eye lens in-vivo
Segmented tomogramSegments: 1000Averaging: 1Tomo. lines: 1000Acq. time: 250 ms
Segmented tomogramSegments: 1000Averaging: 16Tomo. lines: 1000Acq. time: 250 ms
Capsule
Epithelium
Courtesy of Maciej Wojtkowski, Nicolaus Copernicus University, Torun, Poland
M. Szkulmowski,et al. "Efficient reduction of speckle noise in Optical Coherence Tomography," Opt. Express 20, 1337, 2012
CornealTopography
Placido Tomography
Scheimpflugcamera
SS‐OCT
Keratoplasty
Corneal Penetrating Keratoplasty
Anterior surfaceK1 (ax)K2 (ax)
38.38 D (165°)43.88 D (80°)
9.64 mm 35.0 D (146.5°)7.98 mm 42.3 D (56.5°)
9.44 mm 45.8 D (146°)6.82 mm 49.5 D (56°)
Posterior surfaceK1 (ax)
N.A.( )
K2 (ax) 8.39 mm -4.8 D (146.5°)5.73 mm -7.0 (56.5°)
7.03 mm -5.7 D (154.0°)5.29 mm -7.6 D (64.0°)
Pachymetrymin. Thickness
519 µm 572 µm
Courtesy of Maciej Wojtkowski, Nicolaus Copernicus University, Torun, Poland
K. Karnowski, B. J. Kaluzny, M. Szkulmowski, M. Gora, and M. Wojtkowski, Biomedical Opt Express, 2 (9), 2709-2720, 2011
10/29/2012
30
Large scale 3-D imaging
M. Gora et al. Optics Express 17(17), 14880-14894, 2009, I. Grulkowski et al..Opt Express, 17 (6), 4842-4858, 2009.
S. Ortiz et al. Biomed. Opt. Express 2, 3232-3247 (2011) , S. Ortiz et al., Opt Express, 18 (3), 2782-2796, 2010.
K. Karnowski et al., Biomedical Opt Express, 2 (9), 2709-2720, 2011
Courtesy of Maciej Wojtkowski, Nicolaus Copernicus University, Torun, Poland
Conclusions
Future commercial OCT systems will be faster
Accurate topography of optical surfaces Accurate topography of optical surfaces Optical modeling and ray tracing
High quality 3D imaging Contact lens fitting
Implant fitting
3D angle evaluation 3D angle evaluation
Surgery simulation
David Huang, MD, PhD www.COOLLab.net