A deep learning algorithm using CT images to screen for Corona Virus Disease (COVID-19)
Shuai Wang, Bo Kang, Jinlu Ma, Xianjun Zeng, Mingming Xiao, Jia Guo, Mengjiao Cai, Jingyi Yang, Yaodong Li, Xiangfei Meng, Bo Xu
doi: https://doi.org/10.1101/2020.02.14.20023028
Abstract
The outbreak of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-COV-2) has caused approximately 64,000 cases of Corona Virus Disease (COVID-19) in China so far, with that number continuing to grow. To control the spread of the disease, screening large numbers of suspected cases for appropriate quarantine and treatment measures is a priority. Viral nucleic acid testing based on specimens from the lower respiratory tract is the diagnostic gold standard. However, the availability and quality of laboratory testing in the infected region presents a challenge, so alternative diagnostic methods are urgently needed to combat the disease. Based on COVID-19 radiographical changes in CT images, we hypothesized that Artificial Intelligence's deep learning methods might be able to extract COVID-19's specific graphical features and provide a clinical diagnosis ahead of the pathogenic test, thus saving critical time for disease control. To test this possibility, we collected 453 CT images of pathogen-confirmed COVID-19 cases along with previously diagnosed with typical viral pneumonia. 217 images were used as the training set and the inception migration-learning model was used to establish the algorithm. The internal validation achieved a total accuracy of 82.9% with specificity of 80.5% and sensitivity of 84%. The external testing dataset showed a total accuracy of 73.1% with specificity of 67% and sensitivity of 74%. These results indicate the great value of using the deep learning method to extract radiological graphical features for COVID-19 diagnosis.
*注，本文为预印本论文手稿，是未经同行评审的初步报告，其观点仅供科研同行交流，并不是结论性内容，请使用者谨慎使用.
