AI-Powered Multi-Pathology Chest CT Screening System for Rapid Detection of COVID-19 and Other Pulmonary Diseases

Sichuan Online News (Reporter Fan Ruiling): On March 13, West China Hospital of Sichuan University conducted operational testing of the “AI Screening and Auxiliary Diagnostic System for CT Imaging of Multiple Pulmonary Lesions,” with ideal results. The system can rapidly and efficiently analyze patients’ chest CT images to identify imaging features associated with conditions such as pulmonary nodules, lung cancer, tuberculosis, viral pneumonia, and bacterial pneumonia. It can be deployed locally at medical institutions or used remotely via the cloud.

Leveraging the “Big Data Research Platform for Disease Epidemiology,” this project was jointly developed by West China Hospital and Sichuan Xishi Yigou Medical Technology Co., Ltd. in less than 30 days. Built on artificial intelligence deep learning technologies, the project has garnered medical research support from nearly 20 healthcare institutions. The data were derived from clinically confirmed cases, with data selection and annotation performed by radiology imaging experts at West China Hospital.

“CT scanning is a crucial step in screening suspected cases of novel coronavirus pneumonia. However, each patient’s CT scan comprises hundreds of image slices. For physicians, especially those working in epidemic areas, manually reviewing these images one by one places significant demands on visual acuity, physical stamina, and clinical experience.” Li Zhenlin, Director of the Department of Medical Imaging Technology and Deputy Director of the Department of Radiology at West China Hospital, stated that the system’s synchronous analysis of patients’ chest CT images enables the detection of suspected lesions within seconds, thereby reducing physicians’ workload by multiples and saving substantial time for clinical practice. Furthermore, the system can perform quantitative analysis and comparison of chest CT features across different stages of the disease, assisting physicians in predicting disease progression.

It has been learned that, in addition to screening for viral pneumonias including COVID-19, the system can also detect imaging features associated with conditions such as pulmonary nodules, lung cancer, tuberculosis, and bacterial pneumonia.

Following the commercialization of this system, it will hold significant importance for enabling rapid diagnosis of lung CT images in medical institutions and allowing disease control departments to conduct rapid screening of routine lung CT scans from clinical facilities, thereby facilitating the immediate detection of diseases that pose a threat to public health safety. Currently, the system is available free of charge for research use by medical institutions in need, both domestically and internationally.

Special Statement: The views expressed in the article above represent solely those of the author and do not reflect the views or position of Sina.com. If you have any questions regarding the content, copyright, or other matters, please contact Sina.com within 30 days of publication.