A systematic review found that AI is being explored across nearly every area of GI motility diagnostics, from esophageal disorders and gastroparesis to intestinal transit and anorectal dysfunction. The review analyzed 90 studies and found that AI could automate analysis, improve diagnostic accuracy and standardize interpretation, but most models have not yet undergone external validation.
The strongest body of evidence was in esophageal motility, where AI models classified disorders with accuracies of 81%–98% and showed promise in interpreting pH-impedance and FLIP studies. In gastroduodenal disorders, AI was used to analyze gastric emptying and electrogastrography, including models that detected delayed gastric emptying and gastric dysrhythmias. Other applications included wireless capsule endoscopy, intestinal transit, postoperative ileus prediction and anorectal manometry.
Despite the promising results, the researchers emphasized that these technologies remain largely in the research phase. Moving AI into clinical practice will require large-scale prospective validation, demonstrated clinical benefit and regulatory approval.