iCOVID：用于预测COVID-19患者早期恢复时间的可解释深度学习框架。

iCOVID: interpretable deep learning framework for early recovery-time prediction of COVID-19 patients.

作者信息

Wang Jun, Liu Chen, Li Jingwen, Yuan Cheng, Zhang Lichi, Jin Cheng, Xu Jianwei, Wang Yaqi, Wen Yaofeng, Lu Hongbing, Li Biao, Chen Chang, Li Xiangdong, Shen Dinggang, Qian Dahong, Wang Jian

机构信息

School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China.

Department of Radiology, Southwest Hospital, Third Military Medical University (Army Medical University), Chongqing, China.

出版信息

NPJ Digit Med. 2021 Aug 16;4(1):124. doi: 10.1038/s41746-021-00496-3.

Most prior studies focused on developing models for the severity or mortality prediction of COVID-19 patients. However, effective models for recovery-time prediction are still lacking. Here, we present a deep learning solution named iCOVID that can successfully predict the recovery-time of COVID-19 patients based on predefined treatment schemes and heterogeneous multimodal patient information collected within 48 hours after admission. Meanwhile, an interpretable mechanism termed FSR is integrated into iCOVID to reveal the features greatly affecting the prediction of each patient. Data from a total of 3008 patients were collected from three hospitals in Wuhan, China, for large-scale verification. The experiments demonstrate that iCOVID can achieve a time-dependent concordance index of 74.9% (95% CI: 73.6-76.3%) and an average day error of 4.4 days (95% CI: 4.2-4.6 days). Our study reveals that treatment schemes, age, symptoms, comorbidities, and biomarkers are highly related to recovery-time predictions.

大多数先前的研究都集中在开发用于预测COVID-19患者严重程度或死亡率的模型。然而，仍然缺乏用于预测康复时间的有效模型。在此，我们提出了一种名为iCOVID的深度学习解决方案，它可以根据预定义的治疗方案和入院后48小时内收集的异构多模态患者信息，成功预测COVID-19患者的康复时间。同时，一种名为FSR的可解释机制被集成到iCOVID中，以揭示对每个患者预测有重大影响的特征。从中国武汉的三家医院收集了总共3008名患者的数据，用于大规模验证。实验表明，iCOVID可以实现74.9%的时间依赖性一致性指数（95%CI：73.6-76.3%）和4.4天的平均日误差（95%CI：4.2-4.6天）。我们的研究表明，治疗方案、年龄、症状、合并症和生物标志物与康复时间预测高度相关。