Can artificial intelligence standardise colonoscopy quality?
Ce dossier présente deux études chinoises évaluant la performance de systèmes d'aide au diagnostic, basés sur les technologies de l'intelligence artificielle, pour détecter des polypes ou des adénomes colorectaux lors d'une coloscopie
Colonoscopy combined with resection of neoplastic lesions is thought to reduce the incidence of and mortality from colorectal cancer. However, quality of colonoscopy varies among endoscopists. The proportion of patients with one or more adenomas detected during colonoscopy, known as the adenoma detection rate (ADR), is an important indicator of skill and is inversely associated with the incidence of post-colonoscopy colorectal cancer and related mortality. Withdrawal time is postulated as a possible surrogate for mucosal inspection quality and is known to affect the ADR; a withdrawal time of at least 6 min is recommended. Although this inspection time can be technically controlled by the endoscopist, adherence to the recommended time is not necessarily assiduous, even under supervision, and is a major barrier to standardisation of colonoscopies. In The Lancet Gastroenterology & Hepatology, Dexin Gong and colleagues report their idea of adopting artificial intelligence (AI) technology to change this unfavourable situation.
The Lancet Gastroenterology & Hepatology , commentaire, 2019