Home > News Center > Medical Sciences > Sun Yat-sen University Cancer Center Launches Self-Developed, AI-guided Cancer Diagnostic System

Sun Yat-sen University Cancer Center Launches Self-Developed, AI-guided Cancer Diagnostic System

Last updated :2019-10-25

Source: Cancer Center
Written by: Cancer Center
Edited by: Wang Dongmei

Researchers from Sun Yat-sen University Cancer Center have developed an artificial intelligence (AI) medical tool that can help doctors improve the accuracy in detecting upper gastrointestinal cancers through analysis of imaging data from endoscopic examinations. The corresponding research paper has recently been published in the The Lancet Oncology, via the fast-publication track.

Upper gastrointestinal cancers (including esophageal cancer and gastric cancer) are among the most common malignancies and causes of cancer-related deaths worldwide. Most upper gastrointestinal cancers are diagnosed at advanced stages because their signs and symptoms tend to be latent and non-specific. If detected early, the five-year survival rate can exceed 90 percent.

The current method for early detection of the cancers through endoscopy mainly relies on the skills and experience of the physician. However, there is a shortage of health care professionals who are well trained in the use of endoscopes at Chinese county-level hospitals, resulting in a very low rate of early diagnosis.

In this research led by Rui-Hua Xu, President of Sun Yat-sen University Cancer Center, the AI model, named the Gastrointestinal Artificial Intelligence Diagnostic System (GRAIDS), was trained and optimized using more than 1 million endoscopic images from more than 84,000 patients in six hospitals across China. Results showed that the model can diagnose the cancers with an accuracy of over 90 percent.

The GRAIDS achieved high sensitivity in detecting upper gastrointestinal cancers, similar to that of experts who had a minimum of five years of experience in endoscopic procedures, and it has been found to improve the performance of inexperienced endoscopists, which can promote community-based hospitals to improve their performance in upper gastrointestinal cancer diagnoses, said Prof. Rui-Hua Xu.

The research team also developed a cloud-based AI-powered platform for upper gastrointestinal cancer diagnosis. This includes a computer-aided detection (CAD) system allowing for real-time detection of upper gastrointestinal cancer in routine endoscopic examinations; a public website where clinicians and patients can upload endoscopic images and get a second opinion from GRAIDS; and an open-access endoscopic image database for endoscopists training.

Uneven distribution of medical resources between urban and rural areas is one of the major problems in the health care system in China. GRAIDS has been adopted in five hospitals in the country, which can help bridge the gap between tertiary hospitals and primary hospitals in early detection of upper gastrointestinal cancer, said Prof. Rui-Hua Xu.