The Egyptian Journal of Radiology and Nuclear Medicine (2016) 47, 783–792

Egyptian Society of Radiology and Nuclear Medicine

The Egyptian Journal of Radiology andNuclearMedicine

www.elsevier.com/locate/ejrnmwww.sciencedirect.com

ORIGINAL ARTICLE

Comparison of non-attenuation corrected and

attenuation corrected myocardial perfusion SPECT

* Corresponding author at: Atomic Energy Medical Centre (AEMC), Jinnah Post Graduate Medical Centre (JPMC), Rafique Shahee

Karachi, Pakistan. Tel.: +9221 99205693x4; Fax: +9221 99201354.

E-mail addresses: hrnoor@hotmail.com (H. Raza), daringjadoon2002@yahoo.com (L.K. Jadoon).

Peer review under responsibility of The Egyptian Society of Radiology and Nuclear Medicine.

http://dx.doi.org/10.1016/j.ejrnm.2016.05.0060378-603X � 2016 The Egyptian Society of Radiology and Nuclear Medicine. Production and hosting by Elsevier.This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Hasan Raza *, Lubna Khan Jadoon, Sumaira Mushtaq, Aneeqa Jabeen,

Minhaj Maqbool, Musarrat Ul Ain, Bashir Ahmed, Muhammad Ali Memon

Nuclear Medicine Department, Atomic Energy Medical Centre, JPMC, Karachi, Pakistan

Received 3 January 2016; accepted 8 May 2016

Available online 7 June 2016

KEYWORDS

Coronary artery disease;

Myocardial perfusion imag-

ing;

Tc99m sestamibi

Abstract Purpose: To assess the significance of attenuation correction on sensitivity, specificity

and diagnostic accuracy of myocardial perfusion SPECT.

Methods: 102 patients referred for myocardial perfusion imaging (MPI) were divided into two

groups: 42 patients (mean age: 54.6 ± 12.6) were enrolled in the group A, who had undergone coro-

nary artery angiography, within six months of the scan. 60 patients (mean age: 49.79 ± 11.3) were

placed in the group B who had a <15% pretest likelihood of CAD. Both non-corrected (NC) and

attenuation corrected (AC) images were visually analyzed according to 17-segment model of the left

ventricular cavity. Visual assessment derived sensitivity, specificity, diagnostic accuracy and nor-

malcy rate of NC and AC sets of images were compared using McNemar test.

Results: Sensitivity, specificity and diagnostic accuracy for detection of coronary artery disease

were found to be 100%, 11% and 79% respectively for NC images and 66%, 78% and 68% for

AC images. The p value was found to be significant in only the RCA territory. Normalcy rates

in the group B population were 19% for NC image set and 74% for the AC image set. No signif-

icant difference on basis of BMI was observed in attenuation corrected scans. The technique

appears to be more valuable in men.

Conclusion: This study demonstrates that CT based attenuation corrected Tc-99mm sestamibi

SPECT myocardial perfusion imaging significantly improved the specificity of the RCA territory

compared with non-attenuation corrected Tc-99mm sestamibi SPECT myocardial perfusion imag-

ing in both genders irrespective of BMI.� 2016 The Egyptian Society of Radiology and Nuclear Medicine. Production and hosting by Elsevier.

This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-

nd/4.0/).

d Road,

784 H. Raza et al.

1. Introduction

The worth of myocardial perfusion imaging (MPI) is undis-putable. It is not only useful in the diagnosis of coronary

artery disease (CAD) but also helps in risk stratification andmanagement of patients with known or suspected CAD (1–3). Despite the progress in the field of medical technology, soft

tissue attenuation is still a hindrance in the path of the diag-nostic accuracy of myocardial perfusion imaging. Soft tissuephoton attenuation produces inhomogeneous defects thatdecrease the specificity of the test, thereby making it difficult

to distinguish between coronary artery disease and the atten-uation artifact. This may lead to predominant increase in falsepositive studies (4). Diaphragmatic attenuation and breast

attenuation are the most commonly documented artifacts. Inobese persons lateral wall artifacts may also be noted (5).Any or all of these may adversely affect any patient’s study.

Thus it is evident that attenuation artifact is a significantand notorious pitfall of SPECT MPI. Several methods havebeen approved for determining attenuation artifacts such as

prone imaging, electrocardiography (ECG) gated SPECTimaging and attenuation correction (6). For this, the idealway is to create attenuation maps from transmission tomo-graphic image generation through an external source of pho-

tons (7). The external source of photons could be someradionuclide source or X-ray tube. Computed tomographyattenuation correction (CT-AC) has advantages of higher

photon flux and shorter scan times (8).In the current study by using hybrid SPECT/CT (Infinea

Hawkeye), we tried to evaluate does attenuation correction

reduces attenuation artifacts? If yes who are the candidatesin which it could be most usefully employed.

2. Materials and methods

2.1. Study population

A total of 102 patients were prospectively analyzed who werereferred to our department, for either diagnostic or prognosticpurposes as part of their clinical management. They were cat-

egorized into two groups. The group A comprised of 42patients who got coronary angiography done within a periodof six months of cardiac scintigraphy. Their ages ranging from

30 to 74 years (mean age = 54.6 ± 12.6) with male/femaleratio of 36/6. The other group i.e. B composed of 60 patientswho had low likelihood of CAD (<15% pretest probability)

fulfilling Morise criteria (9) and they hadn’t undergone coro-nary angiography. Their ages range from 30 to 75 (meanage = 49.79 ± 11.3) with male/female ratio of 30/30.

The patients selected in the group A did not have MI,LBBB, history of revascularization or cardiomyopathies.These patients underwent coronary angiography 6 monthspre or post MPI SPECT. Coronary angiography was set

as a gold standard. Stenosis of P70% was regarded assignificant.

All patients gave informed consent as part of the proto-

col approved by the ethical committee and have thereforebeen performed in accordance with the ethical standardslaid down in the 1964 Declaration of Helsinki and its later

amendments.

2.2. Stress/rest protocol

Patients were asked to come for cardiac perfusion study after a4 h fasting period and they were advised to refrain from tea,coffee, and chocolate cola drinks at least 12 h before stress.

Beta-blockers were discontinued for 72 h, calcium channelblockers for 48–72 h and long acting nitrates for 12 h beforethe study. Physical examination and baseline ECG were doneand IV line was maintained.

Patients who were unable to exercise in order to achieve tar-get heart rate were infused intravenously with Dipyridamole ata dose of 0.56 mg/kg intravenously over 4-min period. The

patients who were able to exercise underwent symptom limitedexercise tolerance test following Bruce or Modified Bruce pro-tocol to achieve at least 85% of age-adjusted maximal pre-

dicted heart rate and five metabolic equivalents.All the patients were injected with Tc-99m sestamibi follow-

ing One day protocol (10 and 30 mCi for stress and rest imag-

ing respectively).

2.3. Acquisition parameters

Each patient underwent stress/rest Tc-99m MIBI SPECT with

CT-based attenuation correction. Patients were placed in thesupine position with arms over their heads. The SPECT scanswere immediately followed by CT. SPECT images were

acquired using Infinia Hawkeye equipped with two detectorspositioned at 90 degree (L mode).

2.4. Data analysis

All the studies were assessed for any mal-registration andpatient movement. If there was slight misregistration between

emission and transmission data, the images were reprocessedbut discarded if there was evidence of significant patientmotion or misregistration during acquisition.

For visual interpretation of Non-corrected (NC) and CT-

AC views, reconstructed short- and long-axis (horizontal longand vertical long axes) images and polar maps were created byMyovation Perfusion SPECT. The myocardium was divided

into 17 segments as recommended by the quality assurancecommittee of ASNC (10).

All coronary segments were assessed by 2 experienced

nuclear physicians. Stress and rest tomograms were presentedin a random sequence for interpretation to the experiencedobservers. For visual interpretation, no clinical informationwas given to experts, such as patient history and coronary

angiography (CAG) findings. Computer generated myocardialperfusion scores were also masked from the experts. The obser-vers interpreted the both the image sets i.e. NC and AC by rat-

ing each segment as normal or abnormal with a specialemphasis on whether the abnormal segment had reversible ora fixed perfusion defect on short axis and long axis slices as

well as the polar map. A third expert was called in when therewas disagreement between the two experienced nuclearphysicians.

On stress images, a defect in perfusion that was partially orabsolutely resolved in the rest images in two or more contigu-

ous segments or slices was marked as a reversible perfusiondefect. On the other hand, a defect in perfusion on stress

images in two or more contiguous segments or slices that per-

Non-attenuation corrected and attenuation corrected myocardial perfusion SPECT 785

sisted in rest images was called as a fixed perfusion defect. Astudy with fixed or reversible perfusion defect was consideredas an abnormal study. The segments were scored individually

and then in each of the three arterial territories i.e. left anteriordescending (LAD), left circumflex (LCX) and right coronaryartery (RCA).

2.5. Coronary angiography

CAG performed by a standard percutaneous technique which

was reported on visual interpretation of an experiencedangiographer was used for correlation with SPECT findings.Luminal narrowing of P70% was considered significant.

2.6. Statistical analysis

All continuous data such as age had been presented as themean ± standard deviation while categorical data were pre-

sented as percentages. Sensitivity, specificity and diagnosticaccuracy of non-corrected and attenuation corrected imageswere determined in group A. Normalcy rate of non-corrected

and attenuation corrected images was determined in group B.Sensitivity, specificity, diagnostic accuracy and normalcy

rate were compared between attenuation corrected and non-

attenuation corrected sets by use of McNemar test. A p valueof <0.05 was considered significant.

3. Results

In this study, data of 4 patients (1 female and 3 males) in groupA and 7 patients (3 females and 4 males) in group B were dis-

carded due to severe motion artifacts and/or misregistrationbetween CT and SPECT images.

There were 38 patients in group A; their mean age was56.18 ± 12.7 years, with a range of 30–78 years and mean

height was 168.0 ± 25.18 cm (range of 147–185 cm). The studypopulation had mean weight of 70.74 ± 12.25 kg (range of45–90 kg). The range of BMIs was 18–35.8 kg/m2 (Table 1).

Overall, 23 patients (60%), 11 patients (29%) and 4 patients(11%) were categorized as normal weight, overweight andobese respectively in group A.

There were 53 patients in group B with mean age 50.29± 10.5 years with a range of 28–75 years and mean heightwas 156.8 ± 16.5 cm with a range of 140–180 cm. The study

Table 1 Demographics of the study population.

Group A Group B

n 38 53

Age 56.18 ± 12.7 50.29 ± 10.5

BMI (mean ± SD) 24.59 ± 4.07 29.77 ± 6.2

Female 5 (13%) 27 (51%)

Male 33 (87%) 26 (49%)

Hypertension 23 (59%) 32 (61%)

Diabetes 14 (37%) 9 (17%)

Family history 14 (37%) 15 (29%)

Hyperlipidemia 7 (18%) 10 (18%)

Smoking 16 (41%) 12 (19%)

Abnormal BMI 15 (39%) 43 (81%)

Angina 25 (66%) 24 (45%)

population had mean weight of 74.67 ± 15.60 with a rangeof 45–122 kg. The range of BMIs was 19.8–54.2 kg/m2. Thedemographic and clinical characteristics of patients in group

B are shown in Table 1. Overall, 10 patients (19%), 17 patients(32%) and 26 patients (49%) were categorized as normalweight, overweight and obese respectively in group B. A trend

of heavy females was observed in this group.In the group that undergone catheterization (group A),

patients were older, normal weight, were more frequently

men that smoked, and had greater frequencies of diabetes mel-litus, strong family history of CAD, presented commonly withangina and showed equivocal frequency of hyperlipidemia andhypertension compared with low pretest likelihood group.

In the group with a low likelihood of CAD i.e. group B,patients were obese or over weight, relatively younger, morefrequently women, with hypertension as a co-morbid and less

inclined to have hyperlipidemia or diabetes.

3.1. Angiographic characteristics of group A population

Angiographic findings stratified by percentage of luminal nar-rowing P70% and involvement of number of arteries areshown in Table 2. 11 scans were assigned as normal on the

basis of P70% stenosis criterion.

3.1.1. Outcome of the analysis

When the 38 patients in group A were individually analyzed, it

was noted that 15, 26 and 38 number of patients had fixed orreversible perfusion defects on NC images that become 2, 21and 11 on CT-AC images in LAD, LCX and RCA regions

respectively. The number of patients with coronary territorialinvolvement is shown in Table 3.

Similarly, when 53 patients in group B were individuallyevaluated, it was noted that 16, 26 and 38 patients had positive

NC images (i.e. fixed or reversible perfusion defects) out ofwhich only 7, 6 and 4 patients persisted with the perfusiondefects after attenuation correction in LAD, LCX and RCA

regions respectively. The number of patients with coronary ter-ritorial involvement is shown in Table 3.

3.2. Overall sensitivity, specificity, diagnostic accuracy andnormalcy rates

The overall results of 99mTc-sestamibi SPECT in Group A with

and without attenuation correction are depicted in Fig. 1. Theresults revealed that AC improved the overall specificity. How-

Table 2 Angiographic characteristics of group A.

Parameter P70% Stenosis

Normal perfusion 11 (29%)

Single vessel disease 16 (42%)

Double vessel disease 9 (24%)

Triple vessel disease 2 (5%)

LAD region 12 (32%)

LCX region 12 (32%)

RCA region 8 (21%)

LAD indicates left anterior descending artery; LCX, left circumflex

artery; RCA, right coronary artery.

Table 3 Number of patients with perfusion defects according to coronary arterial territories.

Patients with perfusion

defects in LAD territory

Patients with perfusion

defects in LCX territory

Patients with perfusion

defects in RCA territory

Group A NC 15 26 38

AC 2 21 11

Group B NC 16 26 38

AC 7 6 4

Fig. 1 Diagnostic accuracy of MP SPECT in group ‘A’ patients.

Fig. 2 Normalcy rates in group B patients.

Fig. 3 Sensitivity, specificity and diagnostic accuracy of NC and

AC images in LAD region.

Fig. 4 Sensitivity, specificity and diagnostic accuracy of NC and

AC images in LCX region.

Fig. 5 Sensitivity, specificity and diagnostic accuracy of NC and

AC images in RCA region.

786 H. Raza et al.

ever, it is observed that AC reduced the sensitivity and diag-nostic accuracy.

The group B patients also showed a significant increase in

normalcy rates upon attenuation correction. The outcome ofattenuation correction in RCA territory surpassed theimprovement of AC in other coronary regions (Fig. 2).

3.3. Sensitivity, specificity, diagnostic accuracy and normalcyrates according to vascular territory

When all three vascular territories were analyzed indepen-

dently, it came to light that attenuation correction increasedthe specificity in all three territories but it significantlyenhanced in RCA region i.e. from 10% to 79%. The diagnostic

accuracy remained same in LAD and LCX territories butincreased considerably in RCA from 29% to 74%. However,sensitivity declined in all the vascular territories (Figs. 3–5).

In group B, normalcy rates were 72%, 51% and 28% forLAD, LCX and RCA in the non-corrected set of imagesrespectively. The normalcy rate showed significantly highervalues in the attenuation corrected images in all three territo-

ries, more marked in the RCA region. The normalcy ratesfor attenuation corrected images in LAD, LCX and RCAregions were 85%, 89% and 92% respectively (Fig. 2).

Fig. 8 RCA: sensitivity, specificity and diagnostic accuracy of

NC and AC images in female population of group A.

Non-attenuation corrected and attenuation corrected myocardial perfusion SPECT 787

3.4. Influence of gender on the sensitivity, specificity anddiagnostic accuracy of myocardial perfusion scan

Female population of group A comprised of only 5 patients.Due to less number of patients, statistical tests could not be

applied.In the assessment of LAD territory, NC images confirmed

disease in 2 of 2 patients (100% sensitivity) and ruled out itin 2 of 3 patients (66.67% specificity) while AC SPECT verified

LAD disease in 1 of 2 patients (50% sensitivity) and rejected itin 3 of 3 patients (100% specificity). The diagnostic accuracyremained the same in both NC and AC images (Fig. 6).

When evaluating LCX territory NC SPECT had 75% sen-sitivity (3 of 4) that was reduced to 25% (1 of 4) on AC SPECTwhile specificity was discovered to be 100% (1 of 1) on NC

SPECT that remained the same on AC SPECT images. Diag-nostic accuracy that was 80% on NC images fell down to 40%on AC images owing to the decrement in sensitivity (Fig. 7).

In the assessment of RCA territory, it was found that only 1patient had RCA disease on angiography. On NC SPECT,sensitivity was estimated to be 100% as it picked up the diseasethat remained the same on AC images. However, AC enhanced

the specificity from 25% to 100% as it excluded the disease inonly 1 patient on NC images but in 4 in AC images. Diagnosticaccuracy also increased from 40% to 100% when NC and AC

images were compared (Fig. 8).In the male population of group A, it was observed that

attenuation correction improved specificity and diagnostic

accuracy of RCA and LCX territories but substantialenhanced values were noted only in RCA region. LAD terri-

Fig. 6 LAD: sensitivity, specificity and diagnostic accuracy of

NC and AC images in female population of group A.

Fig. 7 LCX: sensitivity, specificity and diagnostic accuracy of

NC and AC images in female population of Group A.

tory did not get benefit from attenuation correction in males(Fig. 9–11). In Fig. 12, myocardial perfusion scan of a 65 yearsold male with 50% blockage in RCA territory of CAD is illus-

Fig. 9 LAD: sensitivity, specificity and diagnostic accuracy of

NC and AC images in male population of group A.

Fig. 10 LCX: sensitivity, specificity and diagnostic accuracy of

NC and AC images in male population of group A.

Fig. 11 RCA: sensitivity, specificity and diagnostic accuracy of

NC and AC images in male population of Group A.

Fig. 12 Short axis (SA), vertical long axis (VLA) and horizontal long axis slices from myocardial perfusion images of a 35-year-old,

hypertensive female with BMI of 23.4 without AC and with AC. Note the reduce uptake in the anterior wall of the myocardium in the

IRNC images due to breast attenuation, which was corrected on the attenuation compensated (IRAC) images.

Fig. 13 Normalcy rates in the female population of group B.

Fig. 14 Normalcy rates in male population of group B.

788 H. Raza et al.

trated, showing fixed perfusion defects in inferior wall in non-corrected images that become normal on attenuation

correction.In group B patients, it was noted that female population

benefitted upon attenuation correction considerably in all the

three vascular regions while the male population largely bene-fitted in the RCA and LCX territories upon attenuation cor-rection with p value <0.05. The conspicuous thing that wasnoted was the reduction of normalcy rate in LAD region in

male population (Figs. 13 and 14). In Fig. 15, myocardial per-fusion scan of a 35 years old female with low probability ofCAD is illustrated, showing fixed perfusion defects in anterior

wall in non-corrected images that become normal on attenua-tion correction.

3.5. Influence of body mass index on the sensitivity, specificityand diagnostic accuracy of myocardial perfusion scan

In the non-obese population of group A, an improvement in

specificity was observed in all the three territories at the costof sensitivity upon attenuation correction. However, signifi-cant improvement (p< 0.05) was noted in RCA region onlykeeping in with the trend noted in both the groups on exclusive

group wise analysis. The population of group B with normalBMI also showed better normalcy rate for the RCA regionas compared to other vascular regions. Therefore, it was noted

that RCA was the fortunate vessel which was getting majoradvantage upon attenuation correction than the other vessels(Table 4).

There were 23 non-obese patients in group A and 2, 6 and15 patients got the advantage in LAD, LCX and RCA regionsrespectively as a result of attenuation correction. Among them,

67% were females that got the edge in LAD, LCX and RCAterritories as 2 out of 3 females showed improved tracer uptakeon attenuation correction in all three territories. There were 10normal weight patients in group B and 1, 2, and 6 patients

showed benefit in LAD, LCX and RCA regions respectivelyon attenuation correction. Among them, 25% were femalesthat showed improvement in each of the three coronary terri-

tories (Table 4).When analyzing RCA territory in group A patients with

BMIP 25, it was found that none of the patients had vesselswith P70% narrowing. All but one patient showed persistent

Fig. 15 Short axis (SA), vertical long axis (VLA) and horizontal long axis slices from myocardial perfusion images of a 65-year-old,

obese, hypertensive man with BMI of 34 without AC and with AC. His angiography revealed 70% obstruction in LAD territory, 40%

impediment in LCX territory and 50% blockage in RCA territory. Attenuation corrected images show normal tracer distribution in

inferior wall that was significantly reduced in the IRNC images due to diaphragmatic attenuation.

Table 4 Effect of BMI on sensitivity, specificity, diagnostic accuracy and normalcy rate.

Coronary region Sensitivity p value Specificity p value Diagnostic accuracy p value Normalcy rate p value

BMI < 25

LAD

NC 64.29 1.0 77.78 1 69.57 1.0 88.9 0.67

AC 57.14 88.89 69.57 100

LCX

NC 90.00 1.0 30.77 0.25 56.52 0.62 77.88 0.45

AC 80.00 53.85 65.22 100.00

RCA

NC 88.89 0.37 14.29 0.008 43.48 0.18 33.33 0.063

AC 55.56 78.57 69.57 88.90

BMI > 25

LAD

NC 57.14 1.0 100 1 80.0 1.0 68.2 0.039

AC 57.14 100 80.0 84.1

LCX

NC 100 0.1 63.64 1.0 73.33 0.5 45.5 <0.01

AC 50.0 63.64 60.0 86.4

RCA

NC 100 06.67 0.001 6.67 0.08 27.3 <0.01

AC 0 80.00 80.00 93.2

Non-attenuation corrected and attenuation corrected myocardial perfusion SPECT 789

perfusion defects on AC images. Therefore, sensitivity couldnot be found in RCA territory. As attenuation correction

excluded disease in 12 patients, specificity increased from just7% to 80%, and diagnostic accuracy increased in the similarmanner. No significant change was noted in LAD or LCX

regions. However, loss in sensitivity was noted in LCXterritory.

In the 15 overweight patients of group A, none benefitted in

LAD territory, only 2 female patients were classified diseasefree after attenuation correction in LCX region and 4 malesshowed improvement in counts in RCA region. Meanwhile

in group B, 35% females showed better tracer uptake inLAD region, 43% females showed improvement in counts in

LCX region and 48% were noted to have increased traceruptake in the RCA region (Table 4).

4. Discussion

Soft tissue attenuation is a very important shortcoming ofSPECT myocardial imaging (5,11). The problem of attenua-

tion effect has considerably been vanquished by attenuationcorrection methods achieved by the technological advance-

790 H. Raza et al.

ment in hardware and new software development. Initially,attenuation correction in areas of non-homogenous attenua-tion such as chest was achieved by radionuclide based trans-

mission images but CT based attenuation correction israpidly emerging as the standard for SPECT (12). Acquisitionby CT based AC method is relatively faster as compared to

radionuclide based method and it also results in less noise.Furthermore, there is no impact of the SPECT radionuclideon CT data and also there is no need to replace decayed trans-

mission source (13). Masood et al. first approved the clinicalworth of CT based AC of MPI and proved that it did indeedimprove the diagnostic accuracy of the MPI SPECT (14).

In our study, we tried to assess the impact of CT based

attenuation correction on myocardial perfusion scintigraphy.We studied scans of patients visually (qualitatively). For visualanalysis, patients were divided into two groups: i.e. patients

who had undergone coronary angiography within six monthsof the myocardial perfusion scan and patients who had notundergone coronary angiography as they had low likelihood

of CAD according to Morise et al’s criteria (9) and werereferred to our institute for myocardial perfusion scan.

4.1. Overall sensitivity, specificity, diagnostic accuracy andnormalcy rates

In our study, we came across significant diminution of perfu-sion defects (both fixed and reversible perfusion defects) in

MPS after attenuation correction as illustrated in Table 3.Thereby, resulting in increased number of myocardial seg-ments been declared to have normal perfusion. Although it

is usually presumed that attenuation artifacts cause fixed per-fusion defects sometimes they may produce reversible perfu-sion defects as well, as was found in our study. The reason

could be due to shift in body and relative anatomic positionsduring stress and rest imaging as stress and rest imaging areacquired separately, resulting in different manifestations of

artifacts on the stress and rest image sets.After initial studies by Ficaro et al. (15,16) there has been

almost a general consensus that attenuation correction withor without scatter correction increases the specificity and nor-

malcy rate in myocardial perfusion SPECT (17–19). Likewisein our study, specificity increased from 11% to 78% (pvalue < 0.05) for ruling out CAD when analyzed, as assessed

in group A patients. The increase in specificity was balanced bythe loss in sensitivity (Fig. 1). The decrease in sensitivity wasdue to the fewer number of pathological segments identified

by the CT–AC images. The loss in sensitivity also resulted infall in diagnostic accuracy. Anyhow, the larger number of seg-ments being detected as true negatives on attenuation correc-tion implied an increased diagnostic yield of CT–AC SPECT

images. Yet, there are studies that have reported significantlyhigher specificity without a trade-off in sensitivity (20–25).Our study also showed considerable improvement in normalcy

rate after attenuation correction i.e. from 19% to 74% (Fig. 2).There are other studies that also showed similar results(4,18,26–35). In patients with low likelihood of CAD the nor-

malcy rate can be obtained as high as 98% (15). The mostprobable reason for comparatively low normalcy rate in ourstudy might be the selection of Morise criteria, as Morise

et al. themselves observed normalcy rate of 85% in their stud-ies (9).

4.2. Sensitivity, specificity, diagnostic accuracy and normalcyrates according to vascular territory

In our study, we noted insignificant increase in specificities ofLAD and LCX territories but profound increment was

observed in areas supplied by RCA (Figs. 3–5). The increasein specificities was accompanied by a decline in sensitivitiesin all vascular territories. Concomitantly, diagnostic accuracywas also increased in RCA region but remained same in other

two vascular regions. We also observed increasing normalcyrates in all vascular areas after attenuation correction in groupB (Fig. 2). Several papers have demonstrated that attenuation

correction improves specificity and normalcy rate without sig-nificant reduction in sensitivity (14,17,18,20,21,24). Harel et al.(27) also testified increment in specificity for the RCA and

LCX territories after employment of attenuation correctionmethod. In this study, the significant increase in specificity inthe RCA region (from 8% in NC images to 79% in AC

images) was accompanied by an insignificant increase in thespecificity of the LAD (from 83% in NC images to 94% inAC images) and LCX (from 42% in NC images to 54% inAC images) regions. The study by Huang et al. also proved

that attenuation correction benefits most to RCA region wherespecificity increased from 77.9% to 98.7% but in LCX regionno significant change was observed. The unusual feature of the

study was reduction in specificity of LAD territory i.e. from94.1 to 82.4% (25).

4.3. Influence of gender on the sensitivity, specificity, diagnostic

accuracy and normalcy rate of myocardial perfusion scan

Regarding influence of gender on AC effects, our studyrevealed some confusing results. In the study, improvement

in specificities was noted in RCA regions in both genders whileinsignificant change was detected in LCx territory in eithergender. However, LAD showed some improvement in female

population. As the female population in group A was verylow, statistical tests could not be applied. But in group Bpatients, normalcy rates improved in all the three vascular

regions in female population while the male population bene-fitted in the LCX and RCA regions only.

4.4. Influence of body mass index on the sensitivity, specificityand diagnostic accuracy of myocardial perfusion scan

In people with abnormal values of BMI (>25), as there ismore soft tissue around the heart and as a consequence, it is

at greater distance from the scintillation camera we assumedthat there would be more attenuation artifacts in obese peopleas compared to non-obese people and that they would get

greater benefit from attenuation correction. Our hypothesishad already been studied and proved by studies conductedby Hansen et al. (36) and Freedman et al. (37) which showed

lower thallium 201 SPECT accuracy in patients with abnormalBMI of >25.

Contrary to the studies by Hansen and Freedman, Thomp-

son and colleagues (38) examined the role of CT–AC in sub-jects with normal and abnormal BMI using cutoff of 30 andfound that attenuation correction proved to be of equal valuein both the groups. The patients with BMI 630 showed

Non-attenuation corrected and attenuation corrected myocardial perfusion SPECT 791

increase in specificity by 18%, whereas for patients with BMI>30 specificity increased to 35% with AC (p = 0.03). How-ever, they did not elaborate their results according to vessel

territories or the segments in which more pronounced effecthad been noted. Shawgi et al studied the effects of attenuationcorrection in subjects with normal body mass and found it as a

valuable tool in this group of patients as well. With attenua-tion correction, they noted improvement in perfusion in54.4% of patients on visual inspection (39).

In our study significant increase in specificity and normalcyrate was detected in RCA region in both non-obese and obesegroups. However, significant increase in normalcy rate wasalso observed in LCX region. On careful assessment, it was

noted that slightly more effect was noted in obese people ascompared to non-obese people (i.e. increase in specificity74% vs 66%, increase in normalcy rate 67% vs. 56%). Metic-

ulous examination of the results showed that there was nomarked impact of attenuation correction in obese vs non-obese people on gender basis.

4.5. Limitations

Our study has certain limitations. First, the study is small in

contrast to high prevalence of CAD. Second, if semi quantita-tive analysis was used for interpretation, the results wouldhave been much improved. Third, no scatter correction wasused; additional scatter subtraction in combination with atten-

uation correction might further improve diagnostic accuracy.Fourth, angiography was taken as reference standard with cut-offs of P70%. 670% stenosis can be hemodynamically signif-

icant, and even patients with normal angiograms could havemicrovascular disease. Fifth, the body contours were obtainedfrom CT mapping, and information regarding breast tissues

was not accurately outlined. The effect of dense breasts inthe anterior wall may be evaluated in further studies.

5. Conclusions

In general, the specificity (normalcy rate) for detection of theabsence of CAD increases by attenuation correction using

Hybrid SPECT/CT and by and large it also significantlychanges the sensitivity for the detection of CAD. For vascularterritory, it markedly improves the specificity (normalcy rate)in RCA region as compared to other vascular regions.

Although this technique is useful in both genders, it provesto be more valuable in men. Subjects with normal body massindex can also show attenuation artifacts and AC is more or

less equally practical in non-obese and obese people.

Conflict of interest

There is no financial or any other relationship which can beconstrued as conflict of interest.

References

(1) Hachamovitch R, Berman DS, Shaw LJ, et al. Incremental

prognostic value of myocardial perfusion single photon emission

computed tomography for the prediction of cardiac death:

differential stratification for risk of cardiac death and myocardial

infarction. Circulation 1998;97:535–43.

(2) Iskander S, Iskandrian AE. Risk assessment using single-photon

emission computed tomographic technetium-99m sestamibi imag-

ing. J Am Coll Cardiol 1998;32:57–62.

(3) Travin MI, Duca MD, Kline GM, et al. Relation of gender to

physician use of test results and to the prognostic value of stress

technetium 99m sestamibi myocardial single-photon emission

computed tomography scintigraphy. Am Heart J 1997;134:73–82.

(4) Taneja S, Mohan SK, Blake GM, Livieratos L, Clarke SEM.

Synergistic impact of attenuation correction and gating in routine

myocardial SPECT reporting: 2 year follow-up study. Nucl Med

Commun 2008;29:390–7.

(5) DePuey EG, Garcia EV. Optimal specificity of thallium-201

SPECT through recognition of imaging artifacts. J Nucl Med

1989;30:441–9.

(6) Singh B, Bateman TM, Case JA, Heller G. Attenuation artifact,

attenuation correction, and the future of myocardial perfusion

SPECT. J Nucl Cardiol 2007;14:153–64.

(7) Bajev MT, Ricart Y, Gasen RA, Mora J, Quintana ME, Benitez

A, et al. Clinical utility of attenuation correction with X-rays in

myocardial perfusion SPECT studies. Rev Esp Med Nucl

2007;26:359–66.

(8) Tonge CM, Manoharan M, Lawson RS, et al. Attenuation

correction of myocardial SPECT studies using low resolution

computed tomography images. Nucl Med Commun

2005;26:231–7.

(9) Morise AP, Haddad WJ, Beckner D. Development and validation

of a clinical score to estimate the probability of coronary artery

disease in men and women presenting with suspected coronary

disease. Am J Med 1997;102:350–6.

(10) Holly TA, Abbott BG, Mallah MA, Calnon DA, Cohen MC,

DiFilippo. Single photon-emission computed tomography. J Nucl

Cardiol 2010:588–94.

(11) DePuey EG. How to detect and avoid myocardial perfusion

SPECT artifacts. J Nucl Med 1994;35:699–702.

(12) Zaidi H, Hasegawa B. Determination of the attenuation map in

emission tomography. J Nucl Med 2003;44:291–315.

(13) Seo Y, Wong KH, Sun M, et al. Correction of photon attenuation

and collimator response for a body-contouring SPECT/CT

imaging system. J Nucl Med 2005;46:868–77.

(14) Masood Y, Liu YH, Depuey G, et al. Clinical validation of

SPECT attenuation correction using X-ray computed tomogra-

phy-derived attenuation maps: multicenter clinical trial with

angiographic correlation. J Nucl Cardiol 2005;12:676–86.

(15) Ficaro EP, Fessler JA, Shreve PD, Kritzman IN, Rose PA,

Corben JR. Simultaneous transmission/emission myocardial per-

fusion tomography. Diagnostic accuracy of attenuation-corrected

Tc-99m sestamibi single-photon emission computed tomography.

Circulation 1996;93:463–73.

(16) Ficaro EP, Fessler JA, Ackermann RI, Rogers WL, Corbett JR,

Schwaiger M. Simultaneous transmission-emission thallium-201

cardiac SPECI’: effect of attenuation correction on myocardial

distribution. J Nucl Med 1995;36:921–31.

(17) Gallowitsch HJ, Sykora J, Mikosch P, et al. Attenuation-

corrected thallium-201 single-photon emission tomography using

a gadolinium-153 moving line source: clinical value and the

impact of attenuation correction on the extent and severity of

perfusion abnormalities. Eur J Nucl Med 1998;25:220–8.

(18) Links JM, Becker LC, Rigo P, Taillefer R, et al. Combined

corrections for attenuation correction, depth dependent blur, and

motion in cardiac SPECT: a multicenter trial. J Nucl Cardiol

2000;7:414–25.

(19) Hendel RC, Berman DS, Cullom SJ, et al. Multicenter clinical

trial to evaluate the efficacy of correction for photon attenuation

and scatter in SPECT myocardial perfusion imaging. Circulation

1999;99:2742–9.

(20) Hendel RC, Corbett JR, Cullom SJ, DePuey EG, Garcia EV,

Bateman TM. The value and practice of attenuation correction

for myocardial perfusion SPECT imaging: a joint position

792 H. Raza et al.

statement from the American Society of Nuclear Cardiology and

the Society of Nuclear Medicine. J Nucl Cardiol 2002;9:135–43.

(21) Links JM, DePuey EG, Taillefer R, Becker LC. Attenuation

correction and gating synergistically improve the diagnostic

accuracy of myocardial perfusion SPECT. J Nucl Cardiol

2002;9:183–7.

(22) Garcia EV. SPECT attenuation correction: an essential tool to

realize nuclear cardiology’s manifest destiny. J Nucl Cardiol

2007;14:16–24.

(23) Bateman TM, James S, Cullom SJ. Attenuation correction single-

photon emission computed tomography myocardial perfusion

Imaging. Semin Nucl Med 2005;35:37–51.

(24) Chen J, Caputlu-Wilson SF, Shi H, Galt JR, Faber TL, Garcia

EV. Automated quality control of emission–transmission

misalignment for attenuation correction in myocardial perfusion

imaging with SPECT-CT systems. J Nucl Cardiol 2006;13:43–9.

(25) Huang R, Li F, Zhao Z, Liu B, Ou X, Tian R, et al. Hybrid

SPECT/CT for attenuation correction of stress myocardial

perfusion imaging. Clin Nucl Med 2011;36(5):344–9.

(26) Vidal R, Buvat I, Darcourt J, et al. Impact of attenuation

correction by simultaneous emission/transmission tomography on

visual assessment of 201 Tl myocardial perfusion images. J Nucl

Med 1999;40:1301–9.

(27) Harel F, Genin R, Daou D, et al. Clinical impact of combination

of scatter, attenuation correction and depth dependent resolution

recovery for 201 Tl studies. J Nucl Med 2001;42:1451–6.

(28) Duvernoy CS, Ficaro EP, Karabajakian MZ, Rose PA, Corbett

JR. Improved detection of left main coronary artery disease

with attenuation-corrected SPECT. J Nucl Cardiol 2000;7:

639–48.

(29) Kluge R, Sattler B, Seese A, Knapp WH. Attenuation correction

by simultaneous emission-transmission myocardial single-photon

emission tomography using a technetium-99m-labelled radio-

tracer: impact on diagnostic accuracy. Eur J Nucl Med

1997;24:1107–14.

(30) Shotwell M, Singh BM, Fortman C, Bauman BD, Lukes J,

Gerson MC. Improved coronary disease detection with quanti-

tative attenuation-corrected Tl-201 images. J Nucl Cardiol

2002;9:52–62.

(31) Prvulovich EM, Lonn AH, Bomanji JB, Jarritt PH, Ell PJ. Effect

of attenuation correction on myocardial thallium-201 distribution

in patients with a low likelihood of coronary artery disease. Eur J

Nucl Med 1997;24:266–75.

(32) Bavelaar-Croon CD, Pauwels EK, Wall VEE. Gated single-

photon emission computed tomographic myocardial imaging: a

new tool in clinical cardiology. Am Heart J 2001;141:383–90.

(33) Nichols K, Tamis J, DePuey EG, Mieres J, Malhotra S, Rozanski

A. Relationship of gated SPECT ventricular function parameters

to angiographic measurements. J Nucl Cardiol 1998;5:295–303.

(34) Fleming RM. A tete-a-tete comparison of ejection fraction and

regional wall motion abnormalities as measured by echocardio-

graphy and gated sestamibi SPECT. Angiology 2002;53:313–21.

(35) Chua T, Kiat H, Germano G, Maurer G, Train VK, Friedman J,

et al. Gated technetium-99m sestamibi for simultaneous assess-

ment of stress myocardial perfusion, postexercise regional ven-

tricular function and myocardial viability. Correlation with

echocardiography and rest thallium-201 scintigraphy. J Am Coll

Cardiol 1994;23:1107–14.

(36) Hansen CL, Woodhouse S, Kramer M. Effect of patient obesity

on the accuracy of thallium-201 myocardial perfusion imaging.

Am J Cardiol 2000;85:749–54.

(37) Freedman N, Schechter D, Klein M, Marciano R, Rozenman Y,

Chisin R. SPECT attenuation artifacts in normal and overweight

persons, insights from a retrospective comparison of Rb-82

positron emission tomography and Tl-201 SPECT myocardial

perfusion imaging. Clin Nucl Med 2000;25:1019–23.

(38) Thompson RC, Heller GV, Johnson LL, et al. Value of

attenuation correction on ECG-gated SPECT myocardial perfu-

sion imaging related to body mass index. J Nucl Cardiol 2005;12

(2):195–202.

(39) Shawgi M, Tonge CM, Lawson RS, Muthu S, James J,

Arumugam P. Attenuation correction of myocardial perfusion

SPET in patients of normal body mass index. Hell J Nucl Med

2012;15(3):215–9.