Welcome to the Tison Laboratory

Combining Complex Multi-Modality Medical Data
With State of the Art Machine Learning
To Deliver AI Driven Insights for Cardiac Precision Medicine

The Tison Lab @ UCSF focuses on using machine learning and large-scale epidemiologic and clinical research methods to achieve goals in disease prevention, prediction and disease phenotyping. We focus on translational projects which have an ability to deliver real-world clinical impact. We draw from various complimentary modalities of data in order to paint a comprehensive and longitudinal picture of real-time patient status and disease phenotype: medical imaging, electronic health record & clinical data, remote sensors (smart phones/devices, remote patient monitoring, meta-data) and population-level epidemiologic data. The Tison Lab is located in the UCSF Division of Cardiology, the Cardiovascular Research Institute and Department of Medicine, and is part of the Bakar Computational Health Sciences Institute and Eureka/Health eHeart. We are part of multi-institutional collaborations including with UC Berkeley and Stanford University.


Tison Lab Announcement Highlights! 

More Announcements in Media

 Invited Article in JAMA Cardiology on leveraging deep learning to expand the diagnostic utility of electrocardiograms (ECG).


 A Nature Reviews Endocrinology article highlights work led by the Tison Lab published in Nature Medicine detecting diabetes with smartphones, discussing the broad potential benefits of its accessibility and some of the work ahead to deploy this biomarker clinically.


   New York Times articles (1 and 2) highlighted Tison Lab work describing worldwide changes in physical activity during COVID-19, using >19 million step count measurements from >455,000 global users.


  A paper led by the Tison Lab was published in Nature Medicine showing how a digital biomarker for diabetes was developed from smartphone-based signals using deep learning. Read more at the project page.


 A paper describing worldwide changes in physical activity since COVID-19 published in the Annals of Internal Medicine.


  A paper published in Nature Medicine in collaboration with Andrew Ng's lab at Stanford pioneered the use of deep learning for electrocardiogram (ECG) analysis.