Presentation 2020-01-25
Fundamental study of extracardiac accumulation using Deep learning in myocardial perfusion SPECT
Yuto Homma, Akihiro kikuchi, Masataka kitama, Keiichi Oda, Koki Okada, Takashi Kawakami,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) [Purpose] The purpose of this study is to evaluate and evaluate digital phantoms using deep learning techniques to remove extracardiac accumulation in myocardial perfusion SPECT. [Methods] Phantom creation and SPECT collection were performed using digital phantom creation software (X-CAT) and software (SIMIND) capable of collecting the data on a PC. From the created projection data, 60 cross-sectional images were created and segmentation of the myocardium, liver, and intestine was performed. These cross-sectional images were trained in a deep learning system as teacher data, extracardiac accumulation was removed by the system, and only the extraction of myocardium was attempted. The evaluation method was to create a Circumferential Profile curve, and the effects of extracardiac accumulation were evaluated. The deep learning system used was developed through joint research. [Results] By learning and evaluating with a digital phantom, extracardiac accumulation was removed and myocardium could be extracted in pixel units. Conclusion: This study suggests that the effects of extracardiac accumulation on myocardial perfusion SPECT examination can be reduced. However, there is a shortage of teacher data and a problem of extraction accuracy due to segmentation.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) myocardial perfusion SPECT / Deep Learning / digital phantom / XCAT / SIMIND / segmentation
Paper # MICT2019-41,MBE2019-70
Date of Issue 2020-01-18 (MICT, MBE)

Conference Information
Committee MBE / MICT
Conference Date 2020/1/25(1days)
Place (in Japanese) (See Japanese page)
Place (in English) KKR Hotel Kumamoto
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair Taishin Nomura(Osaka Univ.) / Shinsuke Hara(Osaka City Univ.)
Vice Chair Takashi Watanabe(Tohoku Univ.) / Eisuke Hanada(Saga Univ.) / Chika Sugimoto(Yokohama National Univ.)
Secretary Takashi Watanabe(Kyushu Univ.) / Eisuke Hanada(Nagoya Inst. of Tech.) / Chika Sugimoto(Kobe Univ.)
Assistant Yasuyuki Suzuki(Osaka Univ.) / Akihiro Karashima(Tohoku Inst. of Tech.) / Takumi Kobayashi(Yokohama National Univ.) / Keita Saku(Kyushu Univ.) / Kai Ishida(NICT) / Kento Takabayashi(Okayama Pref. Univ.)

Paper Information
Registration To Technical Committee on ME and Bio Cybernetics / Technical Committee on Healthcare and Medical Information Communication Technology
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Fundamental study of extracardiac accumulation using Deep learning in myocardial perfusion SPECT
Sub Title (in English) Evaluation using a digital phantom
Keyword(1) myocardial perfusion SPECT
Keyword(2) Deep Learning
Keyword(3) digital phantom
Keyword(4) XCAT
Keyword(5) SIMIND
Keyword(6) segmentation
1st Author's Name Yuto Homma
1st Author's Affiliation Hokkaido University of Science(Hokkaido University of Science)
2nd Author's Name Akihiro kikuchi
2nd Author's Affiliation Hokkaido University of Science(Hokkaido University of Science)
3rd Author's Name Masataka kitama
3rd Author's Affiliation Hokkaido University of Science(Hokkaido University of Science)
4th Author's Name Keiichi Oda
4th Author's Affiliation Hokkaido University of Science(Hokkaido University of Science)
5th Author's Name Koki Okada
5th Author's Affiliation Hokkaido University of Science(Hokkaido University of Science)
6th Author's Name Takashi Kawakami
6th Author's Affiliation Hokkaido University of Science(Hokkaido University of Science)
Date 2020-01-25
Paper # MICT2019-41,MBE2019-70
Volume (vol) vol.119
Number (no) MICT-390,MBE-391
Page pp.pp.11-13(MICT), pp.11-13(MBE),
#Pages 3
Date of Issue 2020-01-18 (MICT, MBE)