Greg Holste

PhD Student @UTAustin @VITAGroupUT | @NSF Graduate Research Fellow
Building deep learning methods to interpret medical images

[Greg Holste @giholste]

@rohan_khera @YaleMed @cards_lab Awesome, congrats!!

[Greg Holste @giholste]

Paper: https://t.co/FqewHPXKL7
Code+Model: https://t.co/4jcG5UwWsA

Explainer thread: https://t.co/u12krMfG3h

Thanks to @JAMA_current @JAMAplusAI and collaborators @kovatti87 @tokmarton!

[Rohan Khera @rohan_khera]

Excited that #PanEcho is now published in @JAMA_current

Kudos to @giholste @ekoikonomou for leading this incredible effort!

Now also validated on ED POCUS studies + RVNet+ (Thx Attila Kovacs & Marton Tokodi for joining us)

See 🧵 below for details!
https://t.co/4dzfzczyg8
#OpenSource #AI

[Greg Holste @giholste]

Awesome to see PanEcho highlighted in a special podcast episode from @JAMA_current!

"This is the sort of thing that if 10 or 15 years ago you said we'd be able to do it, I suspect there would have been a lot of skepticism."

@rohan_khera @ekoikonomou @cards_lab @VITAGroupUT

[JAMA @JAMA_current]

Will 2026 be the year of real-world evidence that #AI can actually improve patient care?

@JAMAplusAI EIC Roy Perlis, MD, MSc, and Associate Editor Yulin Hswen, ScD, MPH, discuss in this special episode marking 1 year since the launch of JAMA+ AI:

https://t.co/ol21UYExUV

[MedAI Group @MedaiStanford]

This Monday Gregory Holste from The University of Texas at Austin will be joining us to talk about their work on AI-enabled echocardiography interpretation. Catch it at 1-2pm PT this Monday on Zoom! Subscribe to https://t.co/Tr8mrytDbz #ML #AI #medicine #healthcare

[Greg Holste @giholste]

Thanks for highlighting our work @DeryaTR_!

#PanEcho is publicly available for research use as well: https://t.co/4jcG5UwoD2.

We hope this can accelerate research on AI for echocardiography and medical AI more broadly

[Derya Unutmaz, MD @DeryaTR_]

Important advance in medical video imaging interpretation through AI:

Echocardiography is a cornerstone of cardiovascular care, but relies on expert interpretation and manual reporting from a series of videos. An AI system, called PanEcho, has been proposed to automate echocardiogram interpretation with multitask deep learning.

This AI system can, potentially accelerate workflows and enable rapid cardiovascular health screening in point-of-care settings with limited access to trained experts.

[Greg Holste @giholste]

@Yuki_Sahashi Thanks @Yuki_Sahashi! Really admire the work you and @David_Ouyang's group are doing

[Yuki Sahashi @Yuki_Sahashi]

心エコー検査の汎用AIがJAMAから報告された。
心エコー検査には30-80程度の動画・静止画があるがこれらを読み込むとレポートを出力する。
何より素晴らしいと感じるのが、中身もさることながら、公開コードの豊富さだろう。
(続く）
https://t.co/IvqCfSDhYX

[Greg Holste @giholste]

@kovatti87 @JAMA_current @rohan_khera @ekoikonomou @tokmarton Pleasure working with you both @kovatti87 @tokmarton !

[Greg Holste @giholste]

@RyanSullyvan @UofMaryland @johnpdickerson @MingCLinCS @furongh @PeterStone_TX @jeffclune @UBC Woooo congrats Ryan!!

[Inside Precision Medicine @Inside_PM]

AI Echo System Has Automated Screening Potential

The #PanEcho deep-learning system offers AI-enabled echo interpretation and may accurately assess the structure and function of the heart
@giholste @utexasece #AI #echocardiography
https://t.co/Rs8DmF9YUO

[Matt Segar @MattSegar]

@RobertAvramMD @rohan_khera @JAMA_current @giholste @ekoikonomou https://t.co/fWiCHi7rtW

JAMA X account posted the non-login-required link

[Greg Holste @giholste]

Thank you @EricTopol for highlighting our work! What an honor.

This is the product of deep interdisciplinary collaboration between engineers & clinicians. Grateful to teammates & mentors @ekoikonomou @rohan_khera

+ Model is open-source for researchers! https://t.co/4jcG5UwWsA

[Eric Topol @EricTopol]

When A.I. can interpret an echocardiogram.
(one of the most complex medical images)
https://t.co/0bQZkmu4jO

[Yale Cardiology @YaleCardiology]

A new study in @JAMA_current led by Yale researchers finds that an AI-enabled tool can interpret echocardiograms with a high degree of accuracy in just a few minutes: https://t.co/wGJBkHMTNZ
@rohan_khera @ekoikonomou @giholste

[JAMA @JAMA_current]

An #AI system that automatically interprets echocardiograms maintained high accuracy across geography and time from complete and limited studies.

@rohan_khera

https://t.co/kltl6Ahs4I

[Greg Holste @giholste]

Vibe coded a new personal website: https://t.co/XrCKww3drV

Do I know how Next.js works? Not in the slightest. Am I using it in my website? Apparently!

Have to say I'm very impressed with @v0, saved me hours and hours

[Yale Internal Medicine @YaleIMed]

An experimental new screening test that pairs AI with portable cardiac ultrasounds could ensure more people receive care for two types of cardiomyopathy, a deadly heart condition.

Learn more from @ekoikonomou and @rohan_khera.

@YaleCardiology @cards_lab
https://t.co/L44CHHsaSO

[Evangelos K. Oikonomou @ekoikonomou]

(1/n) Can #AI-enhanced #POCUS detect under-diagnosed cardiomyopathies?

In @LancetDigitalH, we report our findings using #AI to screen for ATTR-CM and HCM using >90k POCUS videos across the EDs of 2 health systems.

Link: https://t.co/1o1cj0bPm7
@cards_lab @rohan_khera
🧵⬇️

[The Lancet Digital Health @LancetDigitalH]

NEW Research: #AI-guided detection of under-recognised cardiomyopathies on point-of-care cardiac ultrasonography. @ekoikonomou @rohan_khera

Read it here: https://t.co/Xc2N7G7g3D

[Cardiovascular Business @CardioBusiness]

New AI program delivers rapid, accurate echo video assessments @cards_lab @YaleMed #cardiology #AI #imaging https://t.co/xMJZhJ58fO

[Greg Holste @giholste]

Working on #AI for #echofirst? Use PanEcho!

Why PanEcho? The model is trained on >1M echos from all views on 39 diverse tasks, meaning...

- PanEcho is a powerful transfer learner
- It takes *one line* of Python code

See https://t.co/4jcG5UwoD2 for details + reach out with Qs

[Rohan Khera @rohan_khera]

After 2 months of embargo for #AHA24 Late Breaking presentation, we are pleased to announce PanEcho - Complete AI-enabled echocardiography interpretation with multi-task deep learning

Led by @giholste & @ekoikonomou @cards_lab

Preprint, code, model here: https://t.co/lxBuu8NkTD