Echocardiogram Foundation Model -- Application 1: Estimating Ejection Fraction

TL;DR

This paper introduces EchoAI, a foundation model trained on 1.5 million echocardiograms using self-supervised learning, to accurately estimate ejection fraction, matching expert performance and reducing manual effort.

Contribution

The paper presents a novel foundation model for echocardiograms trained with SSL, enabling accurate ejection fraction estimation with minimal labeled data.

Findings

Achieved mean absolute percentage error of 9.40% in ejection fraction estimation.

Model performance comparable to expert sonographers.

Demonstrated effectiveness of SSL on large echocardiogram dataset.

Abstract

Cardiovascular diseases stand as the primary global cause of mortality. Among the various imaging techniques available for visualising the heart and evaluating its function, echocardiograms emerge as the preferred choice due to their safety and low cost. Quantifying cardiac function based on echocardiograms is very laborious, time-consuming and subject to high interoperator variability. In this work, we introduce EchoAI, an echocardiogram foundation model, that is trained using self-supervised learning (SSL) on 1.5 million echocardiograms. We evaluate our approach by fine-tuning EchoAI to estimate the ejection fraction achieving a mean absolute percentage error of 9.40%. This level of accuracy aligns with the performance of expert sonographers.