J Shanghai Jiaotong Univ Sci ›› 2022, Vol. 27 ›› Issue (1): 70-80.doi: 10.1007/s12204-021-2392-3

• Robotics & AI in Interdisciplinary Medicine and Engineering • Previous Articles     Next Articles

Multi-Model Ensemble Deep Learning Method to Diagnose COVID-19 Using Chest Computed Tomography Images

WANG Zhiming1 (王志明), DONG Jingjing2,3∗ (董静静), ZHANG Junpeng1∗ (张军鹏)   

  1. (1. College of Electrical Engineering, Sichuan University, Chengdu 610056, China; 2. Key Laboratory of Aerospace Medicine of Ministry of Education, Air Force Medical University, Xi’an 710032, China; 3. Lintong Rehabilitation and Recuperation Center, PLA Joint Logistic Support Force, Xi’an 710600, China)
  • Received:2021-01-07 Online:2022-01-28 Published:2022-01-14

Abstract: Deep learning based analyses of computed tomography (CT) images contribute to automated diagnosis of COVID-19, and ensemble learning may commonly provide a better solution. Here, we proposed an ensemble learning method that integrates several component neural networks to jointly diagnose COVID-19. Two ensemble strategies are considered: the output scores of all component models that are combined with the weights adjusted adaptively by cost function back propagation; voting strategy. A database containing 8 347 CT slices of COVID- 19, common pneumonia and normal subjects was used as training and testing sets. Results show that the novel method can reach a high accuracy of 99.37% (recall: 0.998 1; precision: 0.989 3), with an increase of about 7% in comparison to single-component models. And the average test accuracy is 95.62% (recall: 0.958 7; precision: 0.955 9), with a corresponding increase of 5.2%. Compared with several latest deep learning models on the identical test set, our method made an accuracy improvement up to 10.88%. The proposed method may be a promising solution for the diagnosis of COVID-19.

Key words: COVID-19, deep learning, computed tomography (CT) images, ensemble model, convolutional neural network

CLC Number: