The system can discover colorectal cancer and precancerous lesions (colon neoplastic polyps) in real time during endoscopic examination. The PC-based system uses AI technology utilizing deep learning for image analysis, high speed processing algorithms, and an advanced image processing device. Using 5,000 endoscopic images from the National Cancer Research Center Central were provided to NEC for AI learning. Using this AI technique resulted in a cancer detection rate of 98%.
NEC news release, July 10, 2017
NEC and National Cancer Research Center develop real-time endoscopic cancer diagnosis based on artificial intelligence