Increase Early Detection + Stage Shift with AI

See how RWJBarnabas Health built a system-wide AI-driven pancreas model, streamlining surveillance to dramatically impact pancreatic cancer outcomes.

37x

increase in pancreatic cyst and cancer patients identified

55%

of cancers identified at non-metastatic stages

77%

return rate for high-risk cyst patients

Download Case Study

“The software can accurately identify a malignancy in or around the pancreas, and more expeditiously move that patient into cancer care pathways.”

Russell C. Langan, MD, RWJBarnabas

THE CHALLENGE

National surveillance remains inconsistent

Pancreatic cysts are the most common identifiable precursor to pancreatic cancer, yet variable referral patterns and manual workflows make follow-up and surveillance difficult.

Key Objectives

Delays in recognition and interval monitoring directly contribute to advanced-stage presentations.

More than half of identified patients do not receive subsequent evaluation.

Most pancreatic cysts are incidentally identified and never linked to appropriate follow-up.

THE SOLUTION

An AI-driven model for system-wide surveillance

Led by Dr. Russell Langan, RWJBarnabas created an automated mechanism to identify abnormalities across 12 hospitals and support evidence-based follow-up at scale.

Key Enablers

New imaging or clinical changes are automatically detected triggering alerts for next steps.

The platform is configured to RWJBarnabas’s preferred criteria for consistent surveillance intervals and escalation.

A Computational Linguistics model detects pancreatic cysts masses and duct changes.

THE RESULT

Unprecedented growth in adherence, early detection

RWJBarnabas identified 37x more patients, generated over 4,600 downstream exams, and identified 55% of cancers at non-metastatic stages—well above historical baselines.

“To improve care quality, and increase early cancer detection at scale, using a powerful AI tool like Eon is essential.”