Studying how AI shapes and how it can serve mental health care.

The AI in Mental Health Research Group (AIMHealth) is a coalition of researchers from the UCSF Department of Psychiatry and Behavioral Sciences and collaborators across UCSF and beyond, investigating large language models, digital psychometrics, EHR phenotyping, autonomous therapeutics, and clinical documentation.

Selected publications

View all →
Integrating Expert Knowledge into Large Language Models Improves Performance for Psychiatric Reasoning and Diagnosis
Sarma, K. V., Hanss, K. E., Halls, A. J. M., Krystal, A., Becker, D. F., Glowinski, A. L., Butte, A. J.
Psychiatry Research · 2025
PDF
Competence or confidence? Assessing the accuracy, reliability, and confidence of large language models in psychiatry
Hanss, K. E., Sarma, K. V., Glowinski, A. L., Krystal, A., Saunders, R., Halls, A. J. M., Gorrell, S., Reilly, E.
JMIR · 2025
PDF PubMed
"You're Not Crazy": A Case of New Onset AI-Associated Psychosis
Pierre, J. M., Gaeta, B., Raghavan, G., Sarma, K. V.
Innovations in Clinical Neuroscience · 2025
PDF

Focus areas

View all →
01

Consumer AI Safety and Digital Mental Health

Prevalence and patterns of generative chatbot use in patients with mental illness.

Read →
02

Digital Psychometrics

LLMs that score validated psychometrics directly from audio & video clinical encounters.

Read →
03

EHR Phenotyping and Knowledge Fusion

Extracting behavioral health signal from unstructured records across institutional boundaries.

Read →
04

Autonomous Diagnostics and Therapeutics

Developing and evaluating AI-driven clinical tools designed for mental health care.

Read →
05

AI Clinical Documentation and Education

Reliability of ambient scribes in psychiatric encounters; ambient AI in resident training.

Read →
06

Research Data Infrastructure

A registry that screens, recruits, and connects patients into AI mental-health research.

Read →