Xiang BaiHanchen WangLiya MaYongchao XuJiefeng GanZiwei FanFan YangKe MaJiehua YangSong BaiChang ShuXinyu ZouRenhao HuangChangzheng ZhangXiaowu LiuDandan TuChuou XuWenqing ZhangXi WangAnguo ChenYu ZengDehua YangMing Wei WangNagaraj HolalkereNeil J. HalinIhab R. KamelJia WuXuehua PengXiang WangJianbo ShaoPattanasak MongkolwatJianjun ZhangWeiyang LiuMichael RobertsZhongzhao TengLucian BeerLorena E. SanchezEvis SalaDaniel L. RubinAdrian WellerJoan LasenbyChuangsheng ZhengJianming WangZhen LiCarola SchönliebTian XiaDepartment of RadiologyDepartment of EngineeringFaculty of MathematicsAlan Turing InstituteStanford University School of MedicineShanghai Institute of Materia Medica, Chinese Academy of SciencesHuazhong University of Science and TechnologyTufts UniversityUniversity of Texas MD Anderson Cancer CenterWuhan University of Science and TechnologyMahidol UniversityStanford UniversityAstraZenecaThe Central Hospital of WuhanThe Johns Hopkins HospitalTongji Medical CollegeMSA CapitalCalmCar IncWuhan Children's HospitalWuhan Blood Centre2022-08-042022-08-042021-12-01Nature Machine Intelligence. Vol.3, No.12 (2021), 1081-1089252258392-s2.0-85121384145https://repository.li.mahidol.ac.th/handle/123456789/76625Artificial 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.Mahidol UniversityComputer ScienceArticleSCOPUS10.1038/s42256-021-00421-z