Deep learning-based digestive tract mucosa staining detection system and method

Abstract

The invention discloses a deep learning-based digestive tract mucosa staining detection system and method. The system comprises an image acquisition device, a host, an intelligent prompt device, a staining device and a staining action capture device, the output end of the image collecting device is connected with the input end of the host, the output end of the host is connected with the input end of the intelligent prompting device, the output end of the intelligent prompting device is connected with the input end of the staining device, and the staining output end of the staining device outputs staining medicine; and the signal output end of the staining device outputs a staining action signal to the staining action capturing device. On the basis of following a clinical guide or expert consensus for screening the early cancer of the digestive tract, the system and method can be suitable for basic medical institutions and most of existing gastrointestinal endoscopes, and can guide, supervise and urge doctors to do digestive endoscopy examination and improve the examination and diagnosis level and efficiency of the doctors.

Images