Deep Learning-Based Recognizing COVID-19 and other Common Infectious Diseases of the Lung by Chest CT Scan Images

Min Fu; Shuang-Lian Yi; Yuanfeng Zeng; Feng Ye; Yuxuan Li; Xuan Dong; Yan-Dan Ren; Linkai Luo; Jin-Shui Pan; Qi Zhang

Based on validation dataset, we evaluated the performance of our AI algorithm in 223 diagnosing the most common infectious diseases of the lung, including 224 non-COVID-19 viral pneumonia, bacterial pneumonia, pulmonary tuberculosis except 225 COVID-19. During training and validation process, accuracy and cross-entropy were 226 plotted against the iteration step, which were shown in Figure 1 . Confusion matrix of 227 the AI framework during validation process was also shown in Figure 1 . Multi-class 228 comparison was performed between COVID-19, non-COVID-19 viral pneumonia, 229 bacterial pneumonia, pulmonary tuberculosis, and normal lung. Binary comparison 230 between COVID-19 and the other four types, including non-COVID-19 viral 231 pneumonia, bacterial pneumonia, pulmonary tuberculosis and normal lung, was also 232 implemented to evaluate the performance of recognizing COVID-19. Accuracy, 233 All rights reserved. No reuse allowed without permission. the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.