Timothy Poterucha

@David_Ouyang @milos_ai Congrats to you and your lab! David, you continue to push the entire field forward. Looking forward to seeing this work in deployment!

1/n We are excited to announce EchoPrime – the first echocardiography AI model capable of evaluating a full transthoracic echocardiogram study, identify the most relevant videos, and produce a comprehensive interpretation! Great work lead by @milos_ai, EchoPrime is the largest and most complex #echofirst AI model yet, trained on 10x the data of existing models.

🚨 We’re hiring a Data Scientist in Cardiology @MayoClinic in Rochester, MN! Work at the cutting edge of AI + cardiology 🫀 PhD or experience in AI required. On-site, high-impact, transformative role. Apply here 👉 jobs.mayoclinic.org/job/rochester/… #AI #Cardiology #DataScience

🧵1/Today, we published a key milestone towards AI based cardiac screening in Nature. doi.org/10.1038/s41586… EchoNext outperformed cardiologists and found thousands of high-risk patients missed in routine care. We also made a version available to the world.

6/6 We're making it public. 📦 EchoNext model weights 📂 100K annotated ECGs 🧪 Code, labels, benchmarks All released to accelerate research. Let’s build the future of cardiovascular screening—together. @PierreEliasMD @Nature @MayoClinic @Columbia #CardioTwitter #AIinMedicine

5/6 📈 We ran EchoNext prospectively on patients with no prior echocardiogram. 🔍 27% were flagged as high-risk. 🩺 In the DISCOVERY trial, EchoNext stratified risk so well that 73% of high-risk patients had undiagnosed SHD. → It works before disease is caught.

1/6 🏥 Structural heart disease (SHD) is common and often missed. What if a standard ECG could spot it—accurately, at scale, across hospitals, and patient types? Meet EchoNext, our AI model trained on 1.2M ECG-echo pairs. 👉 Published in Nature 🔗 doi.org/10.1038/s41586…

The full manuscript is now published open access in @ESC_Journals—dive deeper into methods and results: tinyurl.com/EHJDelineate Let's continue pushing forward AI in cardiovascular care! #Echocardiography #MachineLearning #AIinHealthcare

These findings suggest our AI can help clinicians determine optimal echocardiography follow-up intervals, potentially reducing unnecessary tests and better managing patient care.

🧵 AI for Valvular Heart Disease: DELINEATE-Regurgitation Study Excited to share our latest work on using deep learning to improve diagnosis and prediction in valve regurgitation. Let's dive in! #CardioTwitter #EchoAI @PierreEliasMD tinyurl.com/EHJDelineate (1/8)

We are looking to hire extraordinary data scientists at Mayo Clinic to work on cutting edge cardiac AI problems. Qual: strong technical background (PhD or MS) with experience in medical imaging. On-site, Rochester, MN. Apply at shorturl.at/m9USh or DM.

@David_Ouyang @Yuki_Sahashi @StanfordDeptMed @SmidtHeart @Paul_P_Cheng @milos_ai @rheejust @james_y_zou @bryandhe @ASE360 Extraordinary work, David and team! Way to keep pushing the field forward!

@Yuki_Sahashi @StanfordDeptMed @SmidtHeart @Paul_P_Cheng @milos_ai @rheejust @james_y_zou @bryandhe @ASE360 Preprint: medrxiv.org/content/10.110… GitHub: github.com/echonet/measur… Demo: Below and on Github.

We are thrilled to present EchoNet-Measurements, an open source, comprehensive AI platform for automated #echofirst measurements. Using more than 1,414,709 annotations from 155,215 studies from 78,037 patients for training, this is the most comprehensive #echofirst segmentation model.

@David_Ouyang A tour-de-force. HUGE props to David and team for leading the way on AI echo interpretation!

6/n And as always, we seek to release as much information as possible to advance science. As a resource to the community, we release code and weights for EchoPrime, as well as preprint of our results. EchoPrime: Multi-Video View-Informed Vision-Language Model for Comprehensive Echocardiography Interpretation Preprint: arxiv.org/abs/2410.09704 Model and Weights: github.com/echonet/EchoPr…

8/8 🔍Future work will focus on expanding this to other valvular and cardiac assessments and testing deployment in clinical settings. Hoping this work will contribute to the ongoing revolution in echocardiography and deployment of AI systems to improve cardiovascular health!

7/8 💪The DELINEATE-MR system was validated externally, achieving 79% accuracy and a κ of 0.80. It integrates multiple TTE views, enhancing its diagnostic precision compared to single-view models.

1/8 🧵Excited to share our latest study, published online today in Circulation! Our team at Columbia and Cornell developed the DELINEATE-MR system, an AI-based approach to improve mitral regurgitation (MR) diagnosis by echo. @circAHA @echofirst @pierreeliasMD