AI System Distinguish COVID-19 from Flu and Other Respiratory Diseases in Less Than Three Seconds

A new artificial intelligence (AI) system can help doctors distinguish COVID-19 from flu and other respiratory diseases in less than three seconds.

The AI system developed by scientists at Tsinghua University (Beijing, China) identifies the differences in CT scans of patients to accurately distinguish between these illnesses. Amidst the approaching influenza season and the growing threat of a second wave of COVID-19 infections across the world, the AI system will allow healthcare workers to distinguish between the two respiratory illnesses having similar symptoms.


The researchers developed and evaluated the deep learning-based COVID-19 diagnosis system using multi-class multicenter data, which included 11,356 CT scans from 9,025 subjects consisting of COVID-19, CAP, influenza, and non-pneumonia. CAP subjects included in our database were all nonviral CAP. In such a difficult multi-class diagnosis task, the deep convolutional neural network-based system was able to achieve an area under the receiver operating characteristic curve (AUC) of 97.81% for multi-way classification on test cohort of 3,199 scans, AUC of 92.99% and 93.25% on two publicly available datasets, CC-CCII and MosMedData respectively. In a reader study involving five radiologists, the AI system outperformed all the radiologists in more challenging tasks at a speed of two orders of magnitude above them.

Overall, the AI system was comprehensively validated on large multi-class datasets with higher diagnosis performance than human experts in diagnosing COVID-19. Unlike classical black box deep learning approaches, by visualizing the AI system and applying radiomics analysis, it can decode effective representation of COVID-19 on CT imaging, and potentially lead to the discovery of new biomarkers. Radiologists can perform an individualized diagnosis of COVID-19 with the AI system, adding a new driving force to the fight against the global spread of the outbreak.

Related Links:
Tsinghua University