Resources

BRIM is now available at researchers at UCSF to speed chart abstraction with AI.

Investigators at Johns Hopkins independently evaluated Brim against an established ontology-driven NLP system on 364 unmodified pathology reports spanning pancreatic and breast cancer. Brim averaged 96.7% accuracy across seven cancer registry variables in the pancreatic cohort and 93.7% on breast cancer with no disease-specific tuning, holding up on narrative free-text reports nearly as well as on synoptic ones. The authors position automated extraction as a first-pass layer that pre-populates registry fields and routes ambiguous cases to certified registrars.

Brim Analytics is participating in ARPA-H’s Pediatric Cancer eXpansion (PCX), contributing AI-guided chart abstraction technology to help scale interoperable, research-ready pediatric cancer data infrastructure nationwide.

Brim Analytics appointed Gurmeet Sran, MD, MS, as Physician Advisor to help guide product roadmap development and healthcare market strategy, drawing on his experience as a practicing physician and health system data leader.

Article in the Journal of Medical Systems

An AMIA poster discussing Brim created by our partners at CHOP

The grant extension will expand applications in research, clinical trials, registries and national collaborations.

Brim Analytics announces the launch of its AI-guided chart abstraction tool to streamline medical research.