A new study from the University of California, San Francisco (UCSF) explores how artificial intelligence (AI) can assist in diagnosing hepatorenal syndrome (HRS)—a severe complication of liver disease that remains challenging to identify during hospitalization.

Published in Gastro Hep Advances, the research investigated whether large language models (LLMs) could analyze the clinical notes of multiple healthcare providers to improve diagnostic accuracy and streamline patient management.

“Our approach was inspired by sentiment analysis used in online reviews, where AI summarizes collective opinions,” said Jin Ge, MD, MBA, assistant professor of medicine and gastroenterologist at UCSF, who led the study. “We wanted to see if the same concept could capture the collective clinical judgment about a patient’s condition.”

The UCSF team compared traditional diagnostic methods—which rely on clinical variables such as lab results—with an AI-enhanced model that incorporated sentiment analysis derived from clinicians’ notes. The AI-generated sentiment scores significantly improved predictive accuracy for HRS diagnosis at the time of patient discharge.

According to researchers, the technology helps provide clarity in cases where care teams have differing opinions. By synthesizing diverse clinical perspectives into a unified summary, AI could offer physicians a clearer consensus and support faster, more confident decision-making.

“Using the ‘wisdom of the crowd’ doesn’t just predict outcomes—it reveals what the care team collectively thinks,” Ge explained. “For complex cases with uncertainty, these AI-generated insights could help align decisions and accelerate treatment planning.”

While the model has not yet been applied in clinical settings, researchers see it as a step toward future AI-assisted decision-making tools that could enhance patient outcomes and coordination in hospital environments.

Source: University of California, San Francisco (Gastro Hep Advances Journal)