loading
loading
NIH grant 1IU1RD001795-01: Translational Development of Endoscopic Real Time Computer-Aided Diagnosis of Colorectal Polyps
Institute: {'code': 'VA', 'abbreviation': 'VA', 'name': 'Veterans Affairs', 'admin_org_id': '12500964', 'admin_funding_url': ''}. PI: Eladio Rodriguez-Diaz. Award: $.
Start: 2026-10-01T00:00:00.
Significance to the VA: Colorectal cancer (CRC) is the second leading cause of cancer death in the US, with
nearly 150,000 new cases and 50,000 deaths annually. Colonoscopy with polypectomy remains the gold
standard for CRC screening and surveillance since removal of neoplastic polyps during colonoscopy modifies
disease outcomes and informs subsequent management. Standard practice continues to favor removal of all
visualized polyps for histopathological assessment, despite estimates that nearly half of the polyps are non-
neoplastic, i.e., benign. Studies have shown that the capability to reliably predict polyp pathology endoscopically
in real time could result in substantial improvement in the cost-effectiveness of colonoscopy for CRC
management, something desirable to a capitated healthcare system like the VA. With ~300K colonoscopies
performed at the VA each year, colonoscopy efficiency improvements would be very impactful.
Innovation and Impact: The past decade has seen an explosion in advances in endoscopic technologies toward
diagnosing and treating colorectal neoplasia more precisely. Recent advances in artificial intelligence (AI),
specifically in the field of deep learning, and their application to endoscopic imaging, have shown promise for
enabling real-time endoscopic AI-based optical diagnosis of colorectal polyp pathology, otherwise known as
computer-aided diagnosis (CADx), overcoming operator-based polyp pathology assessment factors such as
interobserver variability, skill, and experience. Such capability would finally open the door to widespread adoption
of cost-saving resect-and-discard and leave-behind paradigms for diminutive polyps, as proposed by the
American Society for Gastrointestinal Endoscopy Preservation and Incorporation of Valuable Endoscopic
Innovations guidelines. More importantly, the incorporation of AI-based quantitative image interpretation into
clinical practice, including in the VA, has the potential to increase early cancer detection and help improve the
care of our Veterans.
Specific Aims: The main objective of our proposal is to further develop, refine, and validate a CADx platform
developed by our group at VA Boston. Specifically, we aim to transition from a research platform to an enterprise-
level system, to enable the purely endoscopic, optical assessment diminutive colorectal polyps (≤ 5mm). The
proposal has two phases/aims: Phase 1: System engineering – where we will update our models and adapt our
existing codebase to run on enterprise hardware; Phase 2: Single site clinical study – where we will conduct a
single center prospective clinical study at VA Boston using the system developed in Phase 1.
Methodology: In Phase 1 we will leverage our existing library of over 1500 recorded colonoscopies to continue
improving our models’ predictive performance and translate our work to run on NVIDIA IGXTM, which is an
industrial-grade, edge AI platform purpose-built for industrial and medical environments. In Phase 2, we will
conduct a clinical study where we expect to enroll approximately 450 patients from the VA Boston CRC screening
program. We will collect both the unassisted and CADx-assisted endoscopist optical diagnosis as well as
standalone CADx predictions and polyp characteristics for polyps found during the procedures. Predictions will
be compared to index pathology to assess performance metrics. These three types of assessment will enable
us to perform comparative analyses and evaluate CADx performance as a standalone diagnostic device versus
as an assistive device.
Path to Translation: We envision a potential deployment path with the GI GeniusTM AI AccessTM platform from
Cosmo Intelligent Medical Devices and Medtronic. This initiative allows third-party developers to deploy their
models/apps using the GI GeniusTM system, which are in use in ~400 endoscopy units across the VA, through
commercial partnerships. By study’s end we expect to have both a GI Genius compatible system and supporting
clinical evidence of its performance to form the basis for FDA interactions.document fetched 2026-10-24 02:46