Multicenter Study Comparing AI-Based Navicam vs. Conventional Pillcam in Small Bowel Pathology

This is a prospective, multicenter, randomized, observational study involving multiple hospitals across Spain. At each site, patients will ingest both the Pillcam SB3 and Navicam capsules in a randomized order. Reading times, transit times, and diagnostic yield will be compared between the two devices. A central reading committee of experienced gastroenterologists will conduct blinded evaluations of both explorations using predefined criteria. The primary endpoint is the diagnostic concordance between Navicam's AI-driven ProScan™ system and the conventional reading of Pillcam SB3, measured by Cohen's kappa index. The secondary endpoints include to assess the correlation in lesion detection, video download times, gastric and small bowel transit times, total reading times, and adverse events. The sample size is 147 patients, accounting for an expected 10% dropout rate, based on previous studies showing a diagnostic concordance kappa index of 0.6. This study aims to establish that the AI-based Navicam capsule is at least as effective as the conventional Pillcam SB3 in diagnosing small bowel lesions, with potentially reduced reading times, thus enhancing clinical efficiency in small bowel diagnostics.