Docbot Announces Publication of UltivisionAI™ to Automatically Assess Bowel Preparation for Colonoscopy in Peer-Reviewed Journal, Gastrointestinal Endoscopy

Docbot Announces Publication of UltivisionAI™ to Automatically Assess Bowel Preparation for Colonoscopy in Peer-Reviewed Journal, Gastrointestinal Endoscopy

Published in Gastrointestinal Endoscopy, Vol. 95, Issue 3

IRVINE, Calif., March 14, 2022 — Docbot, Inc., a leading developer of artificial intelligence (AI)-driven healthcare applications in gastroenterology, announced results from its study evaluating its “UltivisionAI* platform to automate bowel preparation adequacy in patients undergoing colonoscopy procedures. The paper titled “Artificial Intelligence for the Assessment of Bowel Preparation,” was published in a leading journal in its field, Gastrointestinal Endoscopy.

Colonoscopy is considered the standard of care for detection of colonic lesions, but only in a setting of an adequately clean colon. Adequate bowel preparation impacts quality indicators such as adenoma detection rate and cecal intubation rate and is essential for complete inspection of colonic mucosa for pre-cancerous and cancerous lesions.

Through data analyzed retrospectively by Docbot’s UltivisionAI platform, the study showed that a deep learning algorithm can be trained to automatically evaluate bowel cleanliness using the Boston Bowel Preparation Scale (BBPS). When applied to full-length withdrawal-phase colonoscopy videos, the AI algorithm could distinguish adequate from inadequate bowel preparations with high accuracy and assess bowel preparation comparably to 4 gastroenterologists. The high agreement provides promise that a deep learning algorithm may lead to more objective and consistent interpretation of bowel preparation.

Andrew Ritter, CEO of Docbot, commented, “We are excited to announce these peer-reviewed results as Docbot continues to validate a range of product features that can help endoscopists provide the best outcomes for their patients.” William Karnes, M.D., Chief Medical Officer and co-author of the study, added, “Publication by Gastrointestinal Endoscopy validates our clinical AI research by the scientific community. We believe our study is one of the first to demonstrate that convolutional neural networks can be trained to accurately predict the Boston Bowel Prep Score from full-length routine endoscopy videos, rather than curated still images, with high inter-rater agreement with expert human readers.”

Michael Wallace, M.D., Chief of Gastroenterology and Hepatology, Mayo Clinic added, “This study demonstrates the feasibility of AI technology to meet an important unmet need in gastroenterology: automated, accurate and unambiguous quality reporting in colonoscopy.”

About Docbot

Docbot, Inc. develops leading artificial intelligence (AI)-driven healthcare applications in the field of gastroenterology. Its flagship product, UltivisionAI, is a real-time AI image analysis platform for reliable clinical outcomes, practice management workflow efficiencies and data-driven analytics to support the entire practice of GI, including therapeutic areas of colon cancer, esophageal cancer, inflammatory bowel disease screening and diagnosis. The platform can be deployed via endoscopy or capsule endoscopy to empower physicians and provide greater access for patients.


Docbot Media Contact

Jason Garfinkel

*UltivisionAI is an investigational device, and it is not currently cleared by the FDA

SOURCE Docbot, Inc.