News

BCHSI affiliated faculty publish - Alice S. Tang, Sarah R. Woldemariam, Silvia Miramontes, Beau Norgeot, Tomiko T. Oskotsky & Marina Sirota

Real world evidence, AI - Reducing diagnostic delays in acute hepatic porphyria using health records data and machine learning

BCHSI / DoC IT Julia Adler Milstein and Sara Murray (UC Health) lead the new Impact Monitoring Platform for AI in Clinical Care (IMPACC), to robustly assess clinical AI technologies for efficacy, safety, and equity. 

BCHSI Postdoc / First author Chris Williams/UCSF-led study finds artificial intelligence is as good as a physician at prioritizing which patients need to be seen first.

Additional coverage:

• San Francisco Chronicle: Can AI accurately triage ER patients? UCSF study suggests yes, 89% of the time, 5/7/24.
• ScienceDaily: Emergency department packed to the gills? Someday, AI may help, 5/7/24.
• KGO: Using AI to triage ER patients, 5/7/24. (See minute 1:55 for the start of the segment.

affiliated faculty Christopher Y. K. Williams, Travis Zack,  Brenda Y. Miao, et al publish - LLMs may be able to accurately identify higher acuity patient presentation when given pairs of presenting histories extracted from patients’ first ED documentation

Pre-training a generative selective state space model, the Mamba model, on UCSF-specific deidentified clinical notes and time-series structured data - Madhumita Sushil , UCSF
Press release
List of Projects
Whitehouse video

Andreas Rauschecker, MD, PhD, Assistant Professor of Radiology, BCHSI-affiliated faculty, recipient for: Deep Learning Radiogenomic Analysis of Primary CNS Lymphoma on Brain MRI

For cardiologist Rima Arnaout, MD, imaging is the “final frontier” of phenotyping. AI’s ability to analyze images could help echocardiographers like her rule out diseases and abnormalities.

Study by Izzy Freisner, Bakar affiliated faculty Jean Feng, Julian Hong, et al.

Bakar affiliated faculty Julian Hong, et al publish on cost study of their randomized controlled healthcare Machine Learning trial, ML directed supportive care to reduce hospitalizations and ER visits during radiation treatment cost-efficiently

Madhumita Sushil, Travis Zack, Atul Butte, et al publish in NEJM AI on releasing a new oncology NLP benchmark of 40 breast and pancreatic cancer progress notes from UCSF, comprehensively annotated by medical experts.