A Prospective Observational Study to Evaluate the Ability of Artificial Intelligence Navigation System to Recognize Blood Vessels in Submucosal Dissection
Ontology highlight
ABSTRACT: Interventions: After obtaining consent, colorectal endoscopic submucosal dissection (colorectal ESD) is performed according to the usual clinical path (4 days/3 nights) currently in operation at Showa University Northern Hospital. The scope and devices were PCF-H290TI (Olympus Co., Tokyo, Japan), disposable tip attachment (D-201-11304 [Olympus Co., Tokyo, Japan]), and Flush knife Ns 1.5mm (Fujinon Co., Saitama, Japan), exactly the same as those we usually use. The treating physician must be an endoscopist with expertise in ESD (more than 50 cases of upper gastrointestinal ESD and more than 20 cases of colorectal ESD). During ESD, the AI system is operated in real time, but the surgeon does not see the results but only records the output. Therefore, the ESD procedure itself is no different from the usual treatment, and the use of this system should have no impact.
Primary outcome(s): Vessel recognition rate (percentage of frames in which vessels were identified with an IoU index of 0.5 or higher)
Study Design: Single arm Non-randomized
DISEASE(S): Colorectal Neoplasm
PROVIDER: 112942 | ecrin-mdr-crc |
REPOSITORIES: ECRIN MDR
ACCESS DATA