Development and Evaluation of an AI System for COVID-19

Cheng Jin; Weixiang Chen; Yukun Cao; Zhanwei Xu; Xin Zhang; Lei Deng; Chuansheng Zheng; Jie Zhou; Heshui Shi; Jianjiang Feng

Artificial intelligence (AI) may be the unique preparation to take up this challenge. Powered by large labeled datasets [5] and modern GPUs, AI, especially deep learning technique [6] , has achieved excellent performance in several computer vision tasks such as image classification [7] and object detection [8] . Recent research shows that AI algorithms can even achieve or exceed the performance of human experts in certain medical image diagnosis tasks [9] [10] [11] [12] [13] . The AI diagnosis algorithm also has the advantages of high efficiency, high repeatability and easy large-scale deployment. The current outbreak of COVID-19 is worldwide, and the shortage of specialist radiologists threatens the availability and adequacy of screening services for COVID-19 in affected areas. By deploying AI diagnosis algorithms, suspicious patients everywhere, especially in developing countries, will have equal access to right diagnosis, timely treatment and isolation.