STAT July 17, 2019
Casey Ross

Pancreatic cancer often kills people because they are diagnosed too late, after their tumors have spread. Other patients may die following the removal of harmless cysts that appear threatening amid a fog of imaging data and other clinical information.

But a new artificial intelligence system unveiled Wednesday by doctors at Johns Hopkins offers to provide a clearer picture for patients: In testing, it displayed a superhuman ability to differentiate harmful lesions from ones that pose no threat at all.

A study published in Science Translational Medicine detailing the performance of the system, dubbed CompCyst, reported that its use would have reduced 60% of unnecessary surgeries in patients whose tumors were found not to be harmful following their procedures.

“This approach...

Today's Sponsors

LEK
ZeOmega

Today's Sponsor

LEK

 
Topics: AI (Artificial Intelligence), Health System / Hospital, Physician, Provider, Technology
‘Think about the hype’ - AI holds disruptive potential for health care
Will Synthetic, AI-Based Digital Humans Change Pharma and Life Sciences? Q&A with Abid Rahman, SVP Innovation, EVERSANA
Investigators Train AI Systems to Predict RA Outcomes
Confronting the Digital Dilemma in Healthcare’s Quest for Innovation
NIH develops AI tool to better pair cancer patients with drugs

Share This Article