
Looking forward to Co-Chairing this NIH-NHLBI Workshop on Artificial Intelligence for Clinical Decision Support Sept 9-10, 2025 in Bethesda, MD. All are welcome, join us in-person or virtually. Register here: cvent.me/zZlAY1


Looking forward to Co-Chairing this NIH-NHLBI Workshop on Artificial Intelligence for Clinical Decision Support Sept 9-10, 2025 in Bethesda, MD. All are welcome, join us in-person or virtually. Register here: cvent.me/zZlAY1

We analyze performance for ECGs over 4 datasets: PTB-XL, Chapman-Shaoxing, Georgia 12-lead ECG Challenge, and CPSC. @GeoffTison We show you how to visualize the normalized signal for each lead as well as train models with a 7-label task across datasets. arxiv.org/abs/2304.08486

This work shows how #AI can expand the utility of existing medical data and procedures to provide clinically-actionable insights and potentially decrease medical risks. @Jeffolgin, @BarriosEtAl, @Sean_abreau, @DerekSoMD, @Zeeshan_A2

We then externally validated CathEF in a different country @UOHIResearch showing similarly strong performance. This is a technological advancement for angiography, using #AI to assess real-time LV systolic function from any routine angiogram without any added risk or procedures.

We developed and validated CathEF, a video-based neural network, in @UCSFCardiology to estimate continuous LV ejection fraction (LVEF) from standard angiograms of the left coronary artery.

We’re excited to share CathEF in our latest @JAMACardio paper, an #AI algorithm that provides novel information about cardiac pumping function (LVEF) from standard coronary angiograms. jamanetwork.com/journals/jam…. Read more also at tison.ucsf.edu/cath-ef @RobertAvramMD

@GeoffTison @UCSFCardiology created an AI-powered algorithm that accurately predicts 💝myocardial injury using #ECG data. ✅The algorithm's negative predictive value of 80% allows it to triage low-risk patients using a simple and cost-effective ECG. rdcu.be/c6tCu #AI

I am so proud to be part of an amazing community at @UCSF pushing the boundaries of biomedical #AI, including @UCSF_BCHSI @arnaoutlab @GeoffTison @sharmilamajumda @AaronNeinstein @madhumitasushil @UCSF_Ci2 and so many others!

UCSF Center for Real World Evidence Collab w/ @JanssenUS published in @JCardFail "Electrocardiogram Detection of Pulmonary Hypertension Using Deep Learning" sciencedirect.com/science/ar… @atulbutte @GeoffTison RWE, signal processing, AND machine learning / deep learning!

We present a new perspective in @cellrepmed on the present state and future outlook of AI and machine learning in cardiovascular medicine: https://t.co/0xFXtwlSFN @GeoffTison A thread: #AI, #Cardiology

A study published in @LancetGH by lead author @GeoffTison and @UCSF researchers shows that physical activity remains decreased worldwide 2 years since #COVID19 onset. doi.org/10.1016/S2214-109X(2…

AI analysis of HCM ECGs correlates with longitudinal hemodynamic, cardiac structural and laboratory markers in obstructive HCM patients. https://t.co/2JuPKMMkSC #JACC #cvHCM #ECG #EKG #CardioTwitter @GeoffTison @noseworthypeter

Full text link sciencedirect.com/science/ar…

@bmsnews @UCSFHospitals @UCSFCardiology @ucsf_echo @ACCinTouch ucsf.edu/news/2022/03/422396…

AI-ECG may offer a novel paradigm to monitor/track disease status, cardiac hemodynamics and drug therapeutic response. A multi-institutional and industry collaboration with @UCSF @MayoClinic @Myokardia @noseworthypeter @Jeffolgin @drpaulfriedman @sean_abreau @konsiont @zachia5

In the phase 2 PIONEER-OLE study of hypertrophic cardiomyopathy patients receiving Mavacamten drug therapy, our AI-ECG scores tracked with longitudinal trends in intra-cardiac pressures (echo LVOT gradients) and volume overload markers (serum NT-proBNP).