Identification of patients with malignant biliary strictures using a cholangioscopy-based deep learning artificial intelligence (with video).

Researchers

Journal

Modalities

Models

Abstract

Accurate diagnosis of malignant biliary strictures as being benign or malignant remains challenging. Previously, it has been suggested that direct visualization and interpretation of cholangioscopy images has greater accuracy for stricture classification than current sampling techniques (i.e., brush cytology and forceps biopsy) via endoscopic retrograde cholangiopancreatography (ERCP). We aimed to develop a convolutional neural network (CNN) model capable of accurate stricture classification and real-time evaluation based solely on cholangioscopy image analysis.Consecutive patients with cholangioscopy examinations from 2012 to 2021 were reviewed. A CNN was developed and tested using cholangioscopy images with direct expert annotations. The CNN was then applied to a multicenter, reserved test set of cholangioscopy videos. CNN performance was then directly compared to that of ERCP sampling techniques. Occlusion block heatmap analyses were used to evaluate and rank cholangioscopy features associated with malignant biliary strictures.A total of 154 patients with available cholangioscopy examinations were included in the study. The final image database was comprised of 2,388,439 still images. The CNN demonstrated good performance when tasked with mimicking expert annotations of high-quality malignant images (AUROC 0.941). Overall accuracy of CNN-based video analysis (0.906) was significantly greater than that of brush cytology (0.625; p = 0.04) or forceps biopsy (0.609; p =0.03). Occlusion block heatmap analysis demonstrated that the most frequent image feature for a malignant biliary stricture was the presence of frond-like mucosa/papillary projections.This study demonstrates that a CNN developed using cholangioscopy data alone has greater accuracy for biliary stricture classification than traditional ERCP-based sampling techniques.Copyright © 2022 American Society for Gastrointestinal Endoscopy. Published by Elsevier Inc. All rights reserved.