De-identification of 5.3M Radiology Exams to Accelerate AI Training for Fracture Detection at Hospital for Special Surgery (HSS)

Hospital for Special Surgery (HSS) partnered with Dicom Systems to route and de-identify over 5.3 million orthopedic imaging exams to train AI models for fracture detection.

An independent third-party validated full HIPAA Safe Harbor compliance, ensuring zero re-identification risk while preserving longitudinal exam relationships. The result was clean, clinically rich data that accelerated AI model training and deployment.

Challenge: De-identify massive imaging datasets while preserving data utility

Solution: DICOM routing and de-identification via the Unifier® platform

Results:

  • 5.3M radiology exams successfully de-identified
  • Third-party-validated HIPAA Safe Harbor compliance
  • Clinically rich data delivered for AI algorithm development
  • Preserved longitudinal exam relationships

By leveraging Dicom Systems Unifier, HSS provided its AI partner with a fully compliant, high-quality dataset that accelerated model training and improved fracture-detection accuracy.