Advertisement

Accuracy of an artificial neural network for detecting a regional abnormality in myocardial perfusion SPECT

  • Takayuki Shibutani
  • Kenichi Nakajima
  • Hiroshi Wakabayashi
  • Hiroshi Mori
  • Shinro Matsuo
  • Hiroto Yoneyama
  • Takahiro Konishi
  • Koichi Okuda
  • Masahisa Onoguchi
  • Seigo Kinuya
Original Article
  • 56 Downloads

Abstract

Objectives

The patient-based diagnosis with an artificial neural network (ANN) has shown potential utility for the detection of coronary artery disease; however, the region-based accuracy of the detected regions has not been fully evaluated. The aim of this study was to demonstrate the accuracy of all detected regions compared with expert interpretation.

Methods

A total of 109 abnormal regions including 33 regions with stress defects and 76 regions with ischemia were examined, which were derived from 21 patients who underwent myocardial perfusion SPECT within 45 days of coronary angiography. The gray and color scale images, a polar map of stress, rest and difference, and left ventricular function were displayed on the monitor to score the extent and severity of stress defect and ischemia. Two experienced nuclear medicine physicians (Observers A and B) scored the abnormality with a 4-point scale and draw abnormal regions on a polar map. The gold standard was determined by the final judgment of normal or abnormal by the consensus of two other independent expert nuclear cardiologists, and was compared with the stress defect and ischemia derived from ANN.

Results

The concordance rate of ANN to the gold standard was higher than that of two observers. Furthermore, the κ coefficient indicated moderate to substantial agreement for stress defect and slight to the fair agreement for ischemia. The area under the curve (AUC) of ANN was the highest for both stress defect and ischemia; in particular, the ANN of ischemia showed significantly higher AUC than Observer A (p = 0.005). The ANN of stress defect showed higher specificity compared with two observers, while the ANN of ischemia showed higher sensitivity. Consequently, the accuracy of ANN showed the highest in this study.

Conclusion

The ANN-based regional diagnosis showed a high concordance rate with the gold standard and comparable or even higher than the interpretation by nuclear medicine physicians.

Keywords

Artificial neural network Computer-aided diagnosis Myocardial perfusion SPECT Artificial intelligence 

Notes

Compliance with ethical standards

Conflict of interest

K. Nakajima has a collaborative research work with FUJIFILM RI Pharma Co., Ltd., Tokyo, Japan, which developed the cardioREPO software program.

References

  1. 1.
    Underwood SR, Anagnostopoulos C, Cerqueira M, Ell PJ, Flint EJ, Harbinson M, et al. Myocardial perfusion scintigraphy: the evidence. Eur J Nucl Med Mol Imaging. 2004;31:261–91.CrossRefGoogle Scholar
  2. 2.
    Salerno M, Beller GA. Noninvasive assessment of myocardial perfusion. Circ Cardiovasc Imaging. 2009;2:412–24.CrossRefGoogle Scholar
  3. 3.
    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). J Am Coll Cardiol. 2003;42:1318–33.CrossRefGoogle Scholar
  4. 4.
    Golub RJ, Ahlberg AW, McClellan JR, Herman SD, Travin MI, Mather JF, et al. Interpretive reproducibility of stress Tc-99m sestamibi tomographic myocardial perfusion imaging. J Nucl Cardiol. 1999;6:257–69.CrossRefGoogle Scholar
  5. 5.
    Danias PG, Ahlberg AW, Travin MI, Mahr NC, Abreu JE, Marini D, et al. Visual assessment of left ventricular perfusion and function with electrocardiography-gated SPECT has high intraobserver and interobserver reproducibility among experienced nuclear cardiologists and cardiology trainees. J Nucl Cardiol. 2002;9:263–70.CrossRefGoogle Scholar
  6. 6.
    Hachamovitch R, Hayes SW, Friedman JD, Cohen I, Berman DS. Comparison of the short-term survival benefit associated with revascularization compared with medical therapy in patients with no prior coronary artery disease undergoing stress myocardial perfusion single photon emission computed tomography. Circulation. 2003;107:2900–7.CrossRefGoogle Scholar
  7. 7.
    Arsanjani R, Xu Y, Hayes SW, Fish M, Lemley M Jr, Gerlach J, et al. Comparison of fully automated computer analysis and visual scoring for detection of coronary artery disease from myocardial perfusion SPECT in a large population. J Nucl Med. 2013;54:221–8.CrossRefGoogle Scholar
  8. 8.
    Arsanjani R, Xu Y, Dey D, Vahistha V, Shalev A, Nakanishi R, et al. Improved accuracy of myocardial perfusion SPECT for detection of coronary artery disease by machine learning in a large population. J Nucl Cardiol. 2013;20:553–62.CrossRefGoogle Scholar
  9. 9.
    Lomsky M, Richter J, Johansson L, Høilund-Carlsen PF, Edenbrandt L. Validation of a new automated method for analysis of gated-SPECT images. Clin Physiol Funct Imaging. 2006;26:139–45.CrossRefGoogle Scholar
  10. 10.
    Nakajima K, Okuda K, Nyström K, Richter J, Minarik D, Wakabayashi H, et al. Improved quantification of small hearts for gated myocardial perfusion imaging. Eur J Nucl Med Mol Imaging. 2013;40:1163–70.CrossRefGoogle Scholar
  11. 11.
    Nakajima K, Matsuo S, Wakabayashi H, Yokoyama K, Bunko H, Okuda K, et al. Diagnostic performance of artificial neural network for detecting ischemia in myocardial perfusion imaging. Circ J. 2015;79:1549–56.CrossRefGoogle Scholar
  12. 12.
    Nakajima K, Kudo T, Nakata T, Kiso K, Kasai T, Taniguchi Y, et al. Diagnostic accuracy of an artificial neural network compared with statistical quantitation of myocardial perfusion images: a Japanese multicenter study. Eur J Nucl Med Mol Imaging. 2017;44:2280–9.CrossRefGoogle Scholar
  13. 13.
    Nakajima K, Okuda K, Watanabe S, Matsuo S, Kinuya S, Toth K, et al. Artificial neural network retrained to detect myocardial ischemia using a Japanese multicenter database. Ann Nucl Med. 2018.  https://doi.org/10.1007/s12149-018-1247-y (Epub ahead of print).CrossRefPubMedPubMedCentralGoogle Scholar
  14. 14.
    Okuda K, Nakajima K. Normal values and gender differences of left ventricular functional parameters with CardioREPO software. Ann Nucl Cardiol. 2017;3:29–33.CrossRefGoogle Scholar
  15. 15.
    Motwani M, Dey D, Berman DS, Germano G, Achenbach S, Al-Mallah MH, et al. Machine learning for prediction of all-cause mortality in patients with suspected coronary artery disease: a 5-year multicentre prospective registry analysis. Eur Heart J. 2017;38:500–7.PubMedGoogle Scholar
  16. 16.
    Cerqueira MD, Weissman NJ, Dilsizian V, Jacobs AK, Kaul S, Laskey WK, et al. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart. A statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. Circulation. 2002;105:539–42.CrossRefGoogle Scholar
  17. 17.
    Nakajima K, Kumita S, Ishida Y, Momose M, Hashimoto J, Morita K, et al. Creation and characterization of Japanese standards for myocardial perfusion SPECT: database from the Japanese Society of Nuclear Medicine Working Group. Ann Nucl Med. 2007;21:505–11.CrossRefGoogle Scholar
  18. 18.
    Nakajima K, Tamaki N, Kuwabara Y, Kawano M, Matsunari I, Taki J, et al. Prediction of functional recovery after revascularization using quantitative gated myocardial perfusion SPECT: a multi-center cohort study in Japan. Eur J Nucl Med Mol Imaging. 2008;35:2038–48.CrossRefGoogle Scholar
  19. 19.
    Azzarelli S, Galassi AR, Foti R, Mammana C, Musumeci S, Giuffrida G, et al. Accuracy of 99mTc-tetrofosmin myocardial tomography in the evaluation of coronary artery disease. J Nucl Cardiol. 1999;6:183–9.CrossRefGoogle Scholar
  20. 20.
    Nakajima K, Taki J, Higuchi T, Kawano M, Taniguchi M, Maruhashi K, et al. Gated SPET quantification of small hearts: mathematical simulation and clinical application. Eur J Nucl Med. 2000;27:1372–9.CrossRefGoogle Scholar
  21. 21.
    Yoneyama H, Nakajima K, Okuda K, Matsuo S, Onoguchi M, Kinuya S, et al. Reducing the small-heart effect in pediatric gated myocardial perfusion single-photon emission computed tomography. J Nucl Cardiol. 2017;24:1378–88.CrossRefGoogle Scholar
  22. 22.
    Matsumoto N, Berman DS, Kavanagh PB, Gerlach J, Hayes SW, Lewin HC, et al. Quantitative assessment of motion artifacts and validation of a new motion-correction program for myocardial perfusion SPECT. J Nucl Med. 2001;42:687–94.PubMedGoogle Scholar
  23. 23.
    Okuda K, Nakajima K, Kikuchi A, Onoguchi M, Hashimoto M. Cardiac and respiratory motion-induced artifact in myocardial perfusion SPECT. Ann Nucl Cardiol. 2017;3:88–93.CrossRefGoogle Scholar
  24. 24.
    Funahashi M, Shimonagata T, Mihara K, Kashiyama K, Shimizu R, Machida S, et al. Application of pixel truncation to reduce intensity artifacts in myocardial SPECT imaging with Tc-99m tetrofosmin. J Nucl Cardiol. 2002;9:622–31.CrossRefGoogle Scholar
  25. 25.
    van Dongen AJ, van Rijk PP. Minimizing liver, bowel, and gastric activity in myocardial perfusion SPECT. J Nucl Med. 2000;41:1315–7.PubMedGoogle Scholar
  26. 26.
    Hansen CL, Sundaram S. The ratio of the apex/anterior wall: a marker of breast attenuation artifact in women. Nucl Med Commun. 2006;27:803–6.CrossRefGoogle Scholar

Copyright information

© The Japanese Society of Nuclear Medicine 2018

Authors and Affiliations

  • Takayuki Shibutani
    • 1
  • Kenichi Nakajima
    • 2
  • Hiroshi Wakabayashi
    • 2
  • Hiroshi Mori
    • 2
  • Shinro Matsuo
    • 2
  • Hiroto Yoneyama
    • 3
  • Takahiro Konishi
    • 3
  • Koichi Okuda
    • 4
  • Masahisa Onoguchi
    • 1
  • Seigo Kinuya
    • 2
  1. 1.Department of Quantum Medical Technology, Institute of Medical, Pharmaceutical and Health SciencesKanazawa UniversityKanazawaJapan
  2. 2.Department of Nuclear MedicineKanazawa University HospitalKanazawaJapan
  3. 3.Department of Radiological TechnologyKanazawa University HospitalKanazawaJapan
  4. 4.Department of PhysicsKanazawa Medical UniversityKahokuJapan

Personalised recommendations