UMass Chan: AI outperforms biopsy in cancer diagnosis 

Advertisement

Researchers at Worcester-based UMass Chan Medical School tested a real-time AI tool to diagnose cholangiocarcinoma in the first in-human trial of its kind.

The Smart-AI trial, led by Neil Marya, MD, evaluated the tool during live procedures in 41 patients with biliary strictures, according to an April 9 study published in Clinical Gastroenterology and Hepatology.

The AI system analyzed video from a cholangioscope during procedures and demonstrated higher diagnostic accuracy than standard sampling methods, with 87.8% accuracy compared to 67.4%. It also outperformed visual assessments by experienced endoscopists, which had 63.1% accuracy.

Researchers trained the model using prior cholangioscopy videos from UMass Memorial Health and Rochester, Minn.-based Mayo Clinic to identify whether strictures were benign or malignant.

The findings suggest AI could support real-time cancer diagnosis in gastrointestinal care and reduce reliance on traditional biopsy methods.

Advertisement

Next Up in GI & Endoscopy

Advertisement