Self-supervised learning assisted diagnosis for mitral regurgitation severity classification based on color Doppler echocardiography

doi:10.21037/atm-21-3449

	Self-supervised learning assisted diagnosis for mitral regurgitation severity classification based on color Doppler echocardiography
	Yang, Feifei 1,2,3; Zhu, Jiuwen 4; Wang, Junfeng 5; Zhang, Liwei 6; Wang, Wenjun 1; Chen, Xu 1,7; Lin, Xixiang 1,7; Wang, Qiushuang 6; Burkhoff, Daniel 8; Zhou, S. Kevin 4,9,10; He, Kunlun 1,2,3
	2021-10-26
发表期刊	ANNALS OF TRANSLATIONAL MEDICINE
ISSN	2305-5839
页码	14
摘要	Background: Mitral regurgitation (MR) is the most common valve lesion worldwide. However, the quantitative assessment of MR severity based on current guidelines is challenging and time-consuming; strict adherence to applying these guidelines is therefore relatively infrequent. We aimed to develop an automatic, reliable and reproducible artificial intelligence (AI) diagnostic system to assist physicians in grading MR severity based on color video Doppler echocardiography via a self-supervised learning (SSL) algorithm. Methods: We constructed a retrospective cohort of 2,766 consecutive echocardiographic studies of patients with MR diagnosed based on clinical criteria from two hospitals in China. One hundred and forty-eight studies with reference standards were selected in the main analysis and also served as the test set for the AI segmentation model. Five hundred and ninety-two and 148 studies were selected with stratified random sampling as the training and validation datasets, respectively. The self-supervised algorithm captures features and segments the MR jet and left atrium (LA) area, and the output is used to assist physicians in MR severity grading. The diagnostic performance of physicians without and with the support from AI was estimated and compared. Results: The performance of SSL algorithm yielded 89.2% and 85.3% average segmentation dice similarity coefficient (DICE) on the validation and test datasets, which achieved 6.2% and 8.1% improvement compared to Residual U-shape Network (ResNet-UNet), respectively. When physicians were provided the output of algorithm for grading MR severity, the sensitivity increased from 77.0% (95% CI: 70.9-82.1%) to 86.7% (95% CI: 80.3-91.2%) and the specificity was largely unchanged: 91.5% (95% CI: 87.8-94.1%) vs. Conclusions: This study provides a new, practical, accurate, plug-and-play AI-assisted approach for assisting physicians in MR severity grading that can be easily implemented in clinical practice.
关键词	Mitral regurgitation (MR) self-supervised learning (SSL) color Doppler echocardiography mitral regurgitation grading (MR grading)
DOI	10.21037/atm-21-3449
收录类别	SCI
语种	英语
资助项目	Beijing Natural Science and Technology Foundation[7202198] ; Ministry of Industry and Information Technology of China[2020-0103-3-1]
WOS研究方向	Oncology ; Research & Experimental Medicine
WOS类目	Oncology ; Medicine, Research & Experimental
WOS记录号	WOS:000730237900001
出版者	AME PUBL CO
引用统计	被引频次：1[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/18068
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	He, Kunlun
作者单位	1.Chinese Peoples Liberat Army Gen Hosp, Med Big Data Res Ctr, Beijing, Peoples R China 2.Chinese Peoples Liberat Army Gen Hosp, Beijing Key Lab Precis Med Chron Heart Failure, Beijing, Peoples R China 3.Chinese Peoples Liberat Army Gen Hosp, Minist Ind & Informat Technol Biomed Engn & Trans, Key Lab, Beijing, Peoples R China 4.Univ Chinese Acad Sci, Inst Comp Technol, Key Lab Intelligent Informat Proc, MIRACLE Grp, Beijing, Peoples R China 5.Univ Utrecht, Utrecht Inst Pharmaceut Sci, Div Pharmacoepidemiol & Clin Pharmacol, Utrecht, Netherlands 6.Chinese PLA Gen & Lospital, Dept Cardiol, Med Ctr 4, Beijing, Peoples R China 7.China Cardiovasc Res Fdn, Med Sch Chinese PLA, New York, NY USA 8.Cardiovasc Res Fdn, New York, NY USA 9.Univ Sci & Technol China, MTRACLE Ctr, Sch Biomed Engn, Suzhou, Peoples R China 10.Univ Sci & Technol China, Suzhou Inst Adv Res, Suzhou, Peoples R China
推荐引用方式 GB/T 7714	Yang, Feifei,Zhu, Jiuwen,Wang, Junfeng,et al. Self-supervised learning assisted diagnosis for mitral regurgitation severity classification based on color Doppler echocardiography[J]. ANNALS OF TRANSLATIONAL MEDICINE,2021:14.
APA	Yang, Feifei.,Zhu, Jiuwen.,Wang, Junfeng.,Zhang, Liwei.,Wang, Wenjun.,...&He, Kunlun.(2021).Self-supervised learning assisted diagnosis for mitral regurgitation severity classification based on color Doppler echocardiography.ANNALS OF TRANSLATIONAL MEDICINE,14.
MLA	Yang, Feifei,et al."Self-supervised learning assisted diagnosis for mitral regurgitation severity classification based on color Doppler echocardiography".ANNALS OF TRANSLATIONAL MEDICINE (2021):14.