The University of Tokyo · Medicine
Professor Atsuo Yamada's research lab specializes in gastrointestinal endoscopy and artificial intelligence, focusing on improving the diagnosis and management of lower and upper gastrointestinal bleeding using capsule endoscopy and advanced imaging technologies. The lab develops and validates deep learning-based systems to detect critical findings such as colonic diverticular hemorrhage, blood content, and colorectal neoplasms with high accuracy and efficiency. Their work emphasizes clinical applicability, integrating AI tools into real-world endoscopic workflows while considering the expertise level of endoscopists. The lab also conducts randomized controlled trials to evaluate optimal timing and strategies in gastrointestinal bleeding management.
Figures are computed from collected data and may differ slightly.
Hypertension, nonsteroidal anti-inflammatory drugs, and anticoagulants, including aspirin, are independent risk factors for colonic diverticular hemorrhage.
Our CNN-based system for capsule endoscopy videos reduced the reading time of endoscopists without decreasing the detection rate of mucosal breaks. However, the reading level of endoscopists should be considered when using the system.
Abstract Background and Aim Detecting blood content in the gastrointestinal tract is one of the crucial applications of capsule endoscopy (CE). The suspected blood indicator (SBI) is a conventional tool used to automatically tag images depicting possible bleeding in the reading system. We aim to develop a deep learning‐based system to detect blood content in images and compare its performance with that of the SBI. Methods We trained a deep convolutional neural network (CNN) system, using 27 847
In a randomized controlled study, we found that colonoscopy within 24 hours after hospital admission did not increase SRH or reduce rebleeding compared with colonoscopy at 24-96 hours in patients with ALGIB. ClinicalTrials.gov, Numbers: UMIN000021129 and NCT03098173.
Earlier timing of CE achieved a higher diagnostic yield for patients with overt OGIB and consequently resulted in a higher intervention rate.
We developed and validated a new AI-based system that automatically detects colorectal neoplasms in CCE images.
Colonic diverticular bleeding is the most common cause of lower gastrointestinal bleeding. Colonoscopy can be used for both diagnosis and treatment of colonic diverticular bleeding. Identification of the stigmata of recent hemorrhage allows for various endoscopic hemostasis methods. Clipping, endoscopic band ligation, injection therapy, and thermal contact are available methods for endoscopic hemostasis. However, the optimal technique remains to be determined. Herein, we review the techniques an
A new short-type DBE allows faster insertion to the target site for pancreaticobiliary intervention in patients with surgically altered anatomy.
The cumulative incidence of rebleeding at 12 and 60 months was 11.0% and 35.3%, respectively. Five factors enable prediction of not only rebleeding but also transfusion requirements, length of stay, and death in OGIB patients.
The utility of single- and double-balloon endoscopy for colonoscopy seems comparable in patients with incomplete colonoscopy using a conventional colonoscope.
Open papers in the app to read, cite, and organize with AI.
