Case Study: University of Texas Medical Branch (UTMB) achieves AI-ready, shareable imaging datasets with Flywheel

Health System Informatics Leader Uses Flywheel to Create AI-Ready Data Sets

The University of Texas Medical Branch (UTMB), an academic health science center in Galveston, faced a growing challenge: how to organize, share and derive value from massive pathology and radiology imaging assets—urgently needed during the COVID-19 pandemic for remote collaboration and AI research. Director Peter McCaffrey turned to Flywheel and its imaging platform to solve the infrastructure problem of centrally storing, organizing and sharing images across institutions without sending raw files.

Flywheel implemented a “living repository” that automatically ingests images (e.g., Epic-marked suspected COVID x‑rays), de-identifies data, tracks provenance, and supports annotation and analysis so UTMB and partners like UT El Paso can re-run models and reproduce results. The platform reduced internal infrastructure and staffing burdens, streamlined grant submissions, enabled creation of AI-ready, richly annotated datasets for research and commercial collaborations, and made access simple via project controls, CLI and APIs — all measurable operational gains attributed to Flywheel.

University of Texas Medical Branch (UTMB)

Peter McCaffrey

Director of Pathology Informatics and Laboratory Information Systems and Assistant Professor of Pathology