Quantification of myocardial perfusion SPECT using freeware package (cardioBull)
- 218 Downloads
We have developed freeware package for automatically quantifying myocardial perfusion and 123I-labeled radiopharmaceutical single-photon emission computed tomography (SPECT), which is called “cardioBull”. We aim to evaluate diagnostic performance of the detection of coronary artery disease (CAD) on the developed software in comparison with commercially available software package [Quantitative Perfusion SPECT (QPS)].
Stress-rest 99mTc-sestamibi myocardial perfusion SPECT was performed in 36 patients with CAD and 35 control patients. A ≥75% stenosis in the coronary artery was identified by coronary angiography in the CAD group. Segmental perfusion defect score was automatically calculated by both cardioBull and QPS software. Summed stress score (SSS) was obtained to detect CAD by the receiver operator characteristic (ROC) analysis. Areas under the ROC curves (AUC) were calculated in patient-based and coronary-based analyses.
Mean SSSs showed no significant difference between cardioBull and QPS (6.0 ± 7.1 vs. 5.6 ± 7.0). The AUC for cardioBull was equivalent to that for QPS (0.91 ± 0.04 vs. 0.87 ± 0.04, p = n.s.). Sensitivity, specificity, and accuracy for cardioBull were 89, 74, and 82%, respectively. For the regional detection of CAD, the AUC showed largest value in left anterior descending coronary artery (LAD) territory (0.86 ± 0.06 for cardioBull, 0.87 ± 0.06 for QPS, p = n.s.). Sensitivity, specificity and accuracy of cardioBull were 70, 88, and 83% for the LAD; 91, 62, and 66% for the left circumflex coronary artery (LCx); and 78, 69, and 70% for the right coronary artery (RCA), respectively.
The AUC, sensitivity, specificity and accuracy for the detection of CAD showed high diagnostic performance on the developed software. In addition, the developed software provided comparable diagnostic performance to the commercially available software package.
KeywordsMyocardial perfusion SPECT Automatic quantification Coronary artery disease Image processing
We thank Masato Yamada, RT, Minoru Tobisaka, RT, Shigeto Matsuyama, RT, Kojima hironori, RT, Takahiro Konishi, RT (Kanazawa University Hospital, Kanazawa, Japan) and Hidekazu Suzuki, RT (Fussa Hospital, Tokyo, Japan) for their technological assistance. We also thank Dr. Kunihiro Yokoyama (Public Central Hospital of Matto Ishikawa, Ishikawa, Japan) for help with data collection and discussion. The software package was developed as a collaborative work with FUJIFILM RI Pharma Co., Ltd. This work was supported in part by Grants-in-Aid for Scientific Research in Japan (No. 22591320, PI: Kenichi Nakajima).
Conflict of interest
The authors declare no conflict of interests.
- 1.Klocke FJ, Baird MG, Lorell BH, Bateman TM, Messer JV, Berman DS, et al. ACC/AHA/ASNC guidelines for the clinical use of cardiac radionuclide imaging—executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (ACC/AHA/ASNC Committee to Revise the 1995 Guidelines for the Clinical Use of Cardiac Radionuclide Imaging). Circulation. 2003;108:1404–18.PubMedCrossRefGoogle Scholar
- 2.Tamaki N. Guidelines for clinical use of cardiac nuclear medicine (JCS 2005). Circ J. 2005;69(Suppl IV):1125–202.Google Scholar
- 6.Dostbil Z, Ariturk Z, Cil H, Elbey MA, Tekbas E, Yazici M, et al. Comparison of left ventricular functional parameters obtained from three different commercial automated software cardiac quantification program packages and their intraobserver reproducibility. Ann Nucl Med. 2011;25:125–31.PubMedCrossRefGoogle Scholar
- 18.Nakajima K, Okuda K, Kawano M, Matsuo S, Slomka P, Germano G, et al. The importance of population-specific normal database for quantification of myocardial ischemia: comparison between Japanese 360 and 180-degree databases and a US database. J Nucl Cardiol. 2009;16:422–30.PubMedCrossRefGoogle Scholar