Indiana University Southwest Internal Medicine Residency Program Evansville, IN
Muhammad YN. Chaudhary, MBChB1, Inshal Jawed, MBBS2, Muhammad Umair Qadir, MBBS2, Shafaq Jabeen, MD3, Umme Farwa, MD4, Aizaz Anwar Khalid, MBBS5, Oluwagbenga Serrano, MD, FACG6 1Indiana University Southwest Internal Medicine Residency Program, Evansville, IN; 2Dow Medical College, Karachi, Sindh, Pakistan; 3Karachi Medical and Dental College, Karachi, Sindh, Pakistan; 4Jinnah Sindh Medical University, Karachi, Sindh, Pakistan; 5Peshawar Medical College, Karachi, North-West Frontier, Pakistan; 6Indiana University School of Medicine, Vincennes, IN Introduction: Using MCCE, physicians can control a capsule, mainly in the stomach, with greater accuracy for GI tract visualization. MCCE improves on some shortcomings of common wireless capsule endoscopy, and its results are as good as standard endoscopies for many lesions in the GI tract. Integrating artificial intelligence (AI) could boost lesion detection and increase the accuracy of diagnostics in this domain. Methods: The review was done according to the PRISMA guidelines for MCCE and GI disease diagnosis using AI. Publications were gathered from these sources through 2023: PubMed, Embase, and IEEE Xplore. The studies in this systematic review assessed MCCE for detecting gastrointestinal diseases (such as ulcers and neoplasms) and/or applied AI methods to analyze MCCE images. Two team members collected details on how accurate the tests are, the quality of the pictures, and what happens after the scan. We recognized that studies with just a small number of centres might have the risk of bias. Since the outcomes in the studies could not be directly compared, a qualitative synthesis was employed. Results: Among the 348 research papers we reviewed, 15 studies were added. For common GI lesions, MCCE gave results that were as reliable as regular endoscopy; in other words, rates of detecting gastric ulcers and polyps with MCCE were the same as those with endoscopy. Patient positioning and the quality of bowel preparation were important when doing MCCE imaging. AI is mainly used to detect lesions in the small bowel and is now being used to help find lesions in the stomach. AI was shown to help identify lesions in many reports made from MCCE images. Observational research was dominant, and no multiple studies were combined for analysis. Qualitatively, using AI with MCCE led to improved detection of lesions and faster image review. Discussion: AI-assisted MCCE has a high potential to diagnose diseases in the GI system without surgery. It can detect many abnormalities like standard endoscopy, making patients less uncomfortable. Algorithms also alert doctors to any unusual changes, which can help them make their diagnosis more quickly. Current results demonstrate that MCCE technology has advanced greatly, but information on how it detects early cancer and affects long-term outcomes is limited. Further research and bigger studies are essential to establish MCCE+AI and improve how it is applied in practice, but it may help complement other tests.
Disclosures: Muhammad Chaudhary indicated no relevant financial relationships. Inshal Jawed indicated no relevant financial relationships. Muhammad Umair Qadir indicated no relevant financial relationships. Shafaq Jabeen indicated no relevant financial relationships. Umme Farwa indicated no relevant financial relationships. Aizaz Anwar Khalid indicated no relevant financial relationships. Oluwagbenga Serrano indicated no relevant financial relationships.
Muhammad YN. Chaudhary, MBChB1, Inshal Jawed, MBBS2, Muhammad Umair Qadir, MBBS2, Shafaq Jabeen, MD3, Umme Farwa, MD4, Aizaz Anwar Khalid, MBBS5, Oluwagbenga Serrano, MD, FACG6. P2426 - Advancements in Gastrointestinal Diagnostics: A Systematic Review of Magnetic Capsule Endoscopy and Artificial Intelligence Application, ACG 2025 Annual Scientific Meeting Abstracts. Phoenix, AZ: American College of Gastroenterology.
