What is echocardiography?

An ultrasound of the beating heart.

Pump

Function

How strongly the heart squeezes — ejection fraction.

Valves

Flow

Leak and narrowing across all four valves.

Structure

Anatomy

Chambers, wall thickness, and muscle disease.

Bedside

Point-of-care

Portable ultrasound, anywhere a patient is.

The most common cardiac imaging test — millions of studies each year, every one dozens of videos a clinician must review.

Research direction

Cardiac ultrasound for more patients, in more places.

01 · Develop

Precision phenotyping

Spatiotemporal neural networks that measure structure and function from raw echo video.

02 · Understand

Interpretable AI

Foundation models clinicians can inspect, trust, and correct.

03 · Deploy

Trials → implementation

Randomized trials, FDA clearance, and health-system deployment.

Progress, year by year

From one measurement to the whole study.

Increasing the scale of training data · log scale

7,169 videos

Per-model training data — from EchoNet-RV to EchoPrime, across 12 studies (2020–2026).

What our models cover — and the open gaps

A systematic approach for the whole echo exam.

Published / built — click to open On the roadmap

From open science to the clinic

Echo AI that reaches the patient.

Open datasets and models that reproduce expert measurements and reveal subclinical disease — proven in randomized trials, FDA-cleared, and deployed in routine care.

EchoNet
First large echo datasets & code, open-source — a global benchmark (github.com/echonet)
1st
First blinded randomized clinical trial of AI in cardiology (2023)
FDA
510(k)-cleared LVEF; Breakthrough-designated cardiac amyloid
Papers
High-impact results in Nature, NEJM AI, JACC & Circulation

OUYANG LAB · development → deployment → implementation · douyang.github.io

0:00 / 0:00