Senior Associate Professor ZHU Xin(※1) of the University of Aizu and the Department of Coloproctology of Fukushima Medical University Aizu Medical Center jointly developed an AI-based computer-aided detection (CADe) system for colonoscopy (patent pending).
They conducted tests to validate the diagnostic performance of the system and it accurately detected small polyps less than five millimeters in size. The result of the tests was made viewable online on March 16, 2020 and will be shortly published in "Digestive Endoscopy"(※2), the journal of the Japan Gastroenterological Endoscopy Society.
The computer-aided detection (CADe) algorithm was developed by having AI perform deep learning on large numbers of colonoscopy images. The ability of CADe to detect small polyps was almost equivalent to that of veteran colonoscopists.
Please refer to the press release dated on August 31 and the URL below for the details (Japanese text only).
（20200831 PressRelease pdf）
<Comment from Senior Associate Professor Zhu>
I am very glad that we have made such gradual achievements using cutting-edge AI technology in close cooperation with the medical doctors in Fukushima Prefecture. We will work to implement the developed system in continued cooperation with multiple medical institutes in order to contribute to the prevention, early detection, and early treatment for colorectal cancer.