Happy to report that our work on building foundation models for Cardiac MRI has finally been published in Nature Biomedical Engineering! @natBME https://t.co/XKUY0WT0IL
I wrote an app using Google AI Studio that calculates how much oversized or undersized a TAVR valve is based on native valve area or perimeter. I know nothing about coding.
Over the moon to be heading west to Seattle to join the @UWCTSurgRes family!
Grateful to my mentors @karamlou and @Dr_HaniNajm for fostering my passion for cardiothoracic surgery and for the privilege of caring for these patients.
We built Lang1: a 100M–7B family of models specialized for hospital operations.
After finetuning, Lang1-1B outperforms generalist models up to 671B, transfers to unseen tasks and another hospital, and is more data-efficient.
🧵Why small specialists win in healthcare.
#EditorsChoice This study performed left ventricular outflow tract modification in patients with an increased risk of systolic anterior motion. Read the results from Loulmet, Grossi, and coauthors👇
https://t.co/wxKkgvltX6
#VisualAbstract
This review explores the current advancements in xenotransplantation, ongoing challenges, and discuss its future potential in addressing the growing demand for donor organs in cardiac transplantation. @kgrace_phil@NYULangone@PedsLiverTxp@DanielWolbrom https://t.co/F7NWOyxiZs
Congratulations to @stephchangmd for being selected as a Pioneer in healthcare on the #TIME100Health list for her modernization of the double lung transplant. Dr. Chang spoke about her experience during #AATS2025. (Recordings coming soon). https://t.co/e4QLqDagu8
You would expect international STEMI registries to drive quality improvement.
Unfortunately, incorporation bias leads many to mistakenly believe STEMI is a solved problem—yet patients who are missed never even make it into these registries!
Great editorial on this topic by @matt_bivens
A paper in @NatureMedicine reports that large language models can be manipulated to generate misinformation by poisoning a small percentage of data on which they are trained and presents a harm mitigation strategy. https://t.co/CRQwy6jn6R