Advancing COVID-19 diagnosis with privacy-preserving collaboration in artificial intelligence

Paper Info

Title
Advancing COVID-19 diagnosis with privacy-preserving collaboration in artificial intelligence

Abstract
Artificial intelligence provides a promising solution for streamlining COVID-19 diagnoses; however, concerns surrounding security and trustworthiness impede the collection of large-scale representative medical data, posing a considerable challenge for training a well-generalized model in clinical practices. To address this, we launch the Unified CT-COVID AI Diagnostic Initiative (UCADI), where the artificial intelligence (AI) model can be distributedly trained and independently executed at each host institution under a federated learning framework without data sharing. Here we show that our federated learning framework model considerably outperformed all of the local models (with a test sensitivity/specificity of 0.973/0.951 in China and 0.730/0.942 in the United Kingdom), achieving comparable performance with a panel of professional radiologists. We further evaluated the model on the hold-out (collected from another two hospitals without the federated learning framework) and heterogeneous (acquired with contrast materials) data, provided visual explanations for decisions made by the model, and analysed the trade-offs between the model performance and the communication costs in the federated training process. Our study is based on 9,573 chest computed tomography scans from 3,336 patients collected from 23 hospitals located in China and the United Kingdom. Collectively, our work advanced the prospects of utilizing federated learning for privacy-preserving AI in digital health. The COVID-19 pandemic sparked the need for international collaboration in using clinical data for rapid development of diagnosis and treatment methods. But the sensitive nature of medical data requires special care and ideally potentially sensitive data would not leave the organization which collected it. Xiang Bai and colleagues present a privacy-preserving AI framework for CT-based COVID-19 diagnosis and demonstrate it on data from 23 hospitals in China and the United Kingdom.

Year	DOI	Venue
2021	10.1038/s42256-021-00421-z	NATURE MACHINE INTELLIGENCE
DocType	Volume	Issue
Journal	3	12
ISSN	Citations	PageRank
2331-8422	0	0.34
References	Authors
0	46

Authors (46 rows)

Cited by (0 rows)

References (0 rows)

Name	Order	Citations	PageRank
Xiang Bai	1	3517	149.87
Hanchen Wang	2	0	0.34
Liya Ma	3	0	0.34
Yongchao Xu	4	195	14.82
Jiefeng Gan	5	0	0.34
Ziwei Fan	6	0	0.34
Fan Yang	7	0	0.34
Ke Ma	8	98	27.97
Jiehua Yang	9	0	0.68
Song Bai	10	533	33.91
Chang Shu	11	0	0.34
Xinyu Zou	12	0	0.34
Renhao Huang	13	0	0.34
Changzheng Zhang	14	0	0.34
Xiaowu Liu	15	0	0.34
Dandan Tu	16	0	0.34
Chuou Xu	17	0	0.34
Wenqing Zhang	18	0	0.34
Xi Wang	19	47	12.80
Anguo Chen	20	0	0.34
Yu Zeng	21	0	0.34
Dehua Yang	22	0	0.34
Ming-Wei Wang	23	0	0.34
Nagaraj Holalkere	24	0	0.34
Neil J. Halin	25	0	0.34
Ihab R. Kamel	26	0	0.34
Jia Wu	27	0	0.68
Xuehua Peng	28	0	0.34
Xiang Wang	29	0	0.34
Jianbo Shao	30	0	2.70
Pattanasak Mongkolwat	31	0	0.34
Jian J. Zhang	32	80	13.36
Weiyang Liu	33	0	0.34
Michael Roberts	34	0	0.34
Zhongzhao Teng	35	26	5.54
Lucian Beer	36	10	1.10
Lorena Escudero Sanchez	37	0	0.34
Evis Sala	38	9	4.29
Daniel L. Rubin	39	1645	145.14
Adrian Weller	40	141	27.59
Joan Lasenby	41	289	41.28
Chuangsheng Zheng	42	0	0.34
Jianming Wang	43	0	0.34
Z. W. Li	44	24	32.39
Carola-Bibiane Schönlieb	45	334	39.39
Tian Xia	46	0	0.68