P0501 - Comparative Efficacy of SKOUT and GI Genius: A Pilot Study of Artificial Intelligence-Assisted Polyp Detection During Colonoscopy

Carlos Echeverria, MD¹, Jahnavi Udaikumar, MD¹, Simone Jarrett, MD², Maysaa El Zoghbi, MD², Adam Goodman, MD², Aasma Shaukat, MD, MPH, FACG²
1NYU Grossman School of Medicine, Department of Medicine, New York, NY; 2NYU Grossman School of Medicine, Division of Gastroenterology and Hepatology, New York, NY

Introduction: Colorectal cancer (CRC) is a major cause of mortality. Colonoscopy remains the gold standard for prevention, though adenoma detection rates (ADR) vary due to human error and interobserver differences [1]. Computer-aided detection (CADe) systems using artificial intelligence (AI) aim to improve polyp detection and reduce variability. GI Genius (Medtronic) and SKOUT (Iterative Health) are two FDA-approved platforms. GI Genius has demonstrated up to a 30% increase in ADR [2], while SKOUT improves ADR and sessile serrated lesion detection without prolonging withdrawal time [3,6]. Comparative studies are limited. The aim of our study was to compare the real-world performance of both systems across multiple endoscopy centers in our healthcare system.

Methods: We retrospectively analyzed colonoscopies using SKOUT or GI Genius across two endoscopy centers in our healthcare system between June and December 2024. Indications for colonoscopy were classified as either screening, surveillance, or diagnostic. Endoscopic and histopathologic findings were abstracted to assess ADR, polyp detection rate (PDR), adenomas per colonoscopy (APC), adenomas per positive procedure (APP), and true histologic resection (THR). Categorical and continuous variables were compared using the Chi-Squared test or Student’s t-test, respectively.

Results: A total of 233 SKOUT-assisted and 143 GI Genius-assisted colonoscopies were included. As shown in Table 1, ADR for SKOUT vs. GI Genius was 48.1% and 44.1% (p=0.52), respectively. PDR for SKOUT vs. GI Genius was 56.2%, and 60.8% (p=0.44), respectively. The APC for SKOUT vs. GI Genius was 1.14 and 1.08 (p=0.62), respectively. The APP for SKOUT vs. GI Genius was 2.37 and 2.44 (p=0.68), respectively. Lastly, the THR for SKOUT vs. GI Genius was 75.3% and 74.8% (p=0.97), respectively. These findings are also visually represented in Figure 1.

Discussion: Both SKOUT and GI Genius improved adenoma and polyp detection in real-world practice. SKOUT showed slightly higher ADR and APC, while GI Genius had marginally better PDR and APP. No statistically significant differences were observed, supporting comparable efficacy. These findings suggest either platform can enhance colonoscopy quality, with selection based on institutional needs and operator preference. Further prospective studies are needed to clarify potential advantages.


Figure:
PDR = Polyp Detection Rate | APP = Adenomas per Positive Procedure | ADR = Adenoma Detection Rate | APC = Adenomas per Colonoscopy | THR = True Histologic Resection


Figure: PDR = Polyp Detection Rate | APP = Adenomas per Positive Procedure | ADR = Adenoma Detection Rate | APC = Adenomas per Colonoscopy | THR = True Histologic Resection


Disclosures:
Carlos Echeverria indicated no relevant financial relationships.
Jahnavi Udaikumar indicated no relevant financial relationships.
Simone Jarrett indicated no relevant financial relationships.
Maysaa El Zoghbi indicated no relevant financial relationships.
Adam Goodman: Ambu, Inc – Advisor or Review Panel Member. Boston Scientific – Advisor or Review Panel Member. Iterative Health – Advisor or Review Panel Member.
Aasma Shaukat: Freenome inc – Consultant.

Carlos Echeverria, MD¹, Jahnavi Udaikumar, MD¹, Simone Jarrett, MD², Maysaa El Zoghbi, MD², Adam Goodman, MD², Aasma Shaukat, MD, MPH, FACG². P0501 - Comparative Efficacy of SKOUT and GI Genius: A Pilot Study of Artificial Intelligence-Assisted Polyp Detection During Colonoscopy, ACG 2025 Annual Scientific Meeting Abstracts. Phoenix, AZ: American College of Gastroenterology.