Recently, Professor Hu Bing's team of the Department of Gastroenterology of our hospital published a research article Artificial intelligence for detecting and delineating a small flat-type early esophageal squamous cell carcinoma under multimodal in Endoscopy (IF:10.
437), a well-known journal in the field of endoscopy Imaging, revealing the whole process
of the team using the self-developed artificial intelligence-assisted diagnosis system to find a case of early esophageal cancer with the smallest area reported in the international literature and successfully implement endoscopic minimally invasive treatment.
Professor Hu Bing is the corresponding author, and Dr.
Yuan Xianglei is the first author
This paper objectively and comprehensively describes the real-time assisted detection and accurate delineation of this flat early esophageal cancer with an area of only 3mmx2.
6mm in four endoscopic modes (white light endoscopy, iodine staining, non-magnified and magnified narrowband spectral imaging
After the early esophageal cancer was found, the patient immediately received endoscopic minimally invasive treatment, and the postoperative pathological examination results confirmed that the tumor invaded the mucosal lamina propria, which belonged to the T1A stage and had a good
For such a small early esophageal cancer lesion, as long as the doctor is not careful or moves a little faster, it is easy to miss the diagnosis
In recent years, artificial intelligence (AI) assisted diagnosis technology has developed rapidly and made great progress
in early esophageal cancer.
However, most AI-assisted diagnostic products for early esophageal cancer can only roughly mark the suspicious lesion site in the form of a rectangular box, and cannot accurately outline the lesion boundary
in the image.
Professor Hu Bing's team focused on the pain points and difficulties of similar products and developed this AI-assisted diagnosis product
that can perform real-time auxiliary detection in multiple endoscopic modes and outline the boundary of early esophageal cancer.
This AI-assisted diagnosis system is not only the world's first artificial intelligence system that combines multiple endoscopic modes for the detection of early esophageal cancer, but also the world's first artificial intelligence system
that can accurately delineate the boundary of early esophageal cancer under white light endoscopy.
The early esophageal cancer AI-assisted diagnosis product developed by Professor Hu Bing's team is suitable for the mainstream digestive endoscopy instruments currently used in most hospitals, and can be "embedded" in endoscopic displays, which is convenient to use on a single screen, without changing the operating habits of endoscopists, and has high clinical practicability
The product has obtained the Class II medical device certificate and was shortlisted for the 2022 Ministry of Industry and Information Technology Artificial Intelligence Medical Device Innovation Task Unveiling
The successful development and clinical application of AI-assisted diagnosis products for early esophageal cancer is not only conducive to improving the detection rate of early esophageal cancer in grassroots hospitals in China, realizing early detection, early diagnosis and early screening, and improving the prognosis of esophageal cancer patients, but also has great significance
for promoting technological innovation and transformation in key areas of esophageal cancer diagnosis.