Accuracy of artificial intelligence–assisted detection of upper GI lesions

Post written by Wai K. Leung, MD, FRCP, from the Department of Medicine, Queen Mary Hospital, University of Hong Kong, Hong Kong.

Leung_headshot

This meta-analysis aims to summarize the performance of all recently published studies on the use of artificial intelligence (AI)-assisted endoscopic detection of gastric and esophageal neoplastic lesions as well as H. pylori infection.

With the rapid development on AI-assisted medical diagnostics, there is a pressing need to rationalize the latest data on the performance of AI-assisted endoscopic detection of upper-gastrointestinal (GI) lesions. Unlike detection of colonic pathology, which is rather restrictive to polyp detection, there are multiple upper-GI pathologies in the esophagus and stomach that would require independent analysis. 

In this meta-analysis of 23 studies, including more than 960,000 endoscopic images, we found that the accuracy of AI detection, expressed as area under the hierarchical summary receiver-operating characteristic curve (AUC), of neoplastic lesions in the stomach, Barrett’s esophagus, and squamous esophagus and HP status were 0.96 (95% confidence interval [CI], 0.94-0.99), 0.96 (95% CI, 0.93-0.99), 0.88 (95% CI, 0.82-0.96), and 0.92 (95% CI, 0.88-0.97), respectively. The performance of the AI was also superior to endoscopists in the detection of neoplastic lesions in the stomach (AUC, 0.98 vs 0.87; P < 0.001), Barrett’s esophagus (AUC, 0.96 vs 0.82; P < .001), and HP status (AUC, 0.90 vs 0.82; P < 0.001).

Despite the promising performance of early studies on AI-assisted detection of upper-GI lesions, these results were largely based on retrospective reviews of selected images only. Moreover, most AI platforms were trained on detection of a single pathology rather than all upper-GI pathologies. Further multi-center prospective randomized studies are needed to verify the actual benefits of AI-assisted endoscopic detection of upper-GI neoplastic lesions.

Leung

Figure 1. Diagnostic performance of artificial intelligence on detection of neoplastic lesions in the stomach. TP, True positive; FP, false positive; TN, true negative; FN, false negative; AUROC, area under the receiver-operating characteristic curve.

Read the full article online.

The information presented in Endoscopedia reflects the opinions of the authors and does not represent the position of the American Society for Gastrointestinal Endoscopy (ASGE). ASGE expressly disclaims any warranties or guarantees, expressed or implied, and is not liable for damages of any kind in connection with the material, information, or procedures set forth.

Leave a Comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s