Cardiac Output Prediction from Echocardiograms: Self-Supervised Learning with Limited Data

TL;DR

This paper introduces a self-supervised learning approach using SimCLR to enhance cardiac output prediction from echocardiograms, especially effective with limited data, reducing overfitting and outperforming large-scale models.

Contribution

It presents a novel SSL pretraining strategy for echocardiogram-based CO prediction that performs well with limited datasets, improving model robustness and accuracy.

Findings

SSL improves representation learning in limited data scenarios.

Achieves Pearson correlation of 0.41 on test set.

Outperforms a model trained on over one million exams.

Abstract

Cardiac Output (CO) is a key parameter in the diagnosis and management of cardiovascular diseases. However, its accurate measurement requires right-heart catheterization, an invasive and time-consuming procedure, motivating the development of reliable non-invasive alternatives using echocardiography. In this work, we propose a self-supervised learning (SSL) pretraining strategy based on SimCLR to improve CO prediction from apical four-chamber echocardiographic videos. The pretraining is performed using the same limited dataset available for the downstream task, demonstrating the potential of SSL even under data scarcity. Our results show that SSL mitigates overfitting and improves representation learning, achieving an average Pearson correlation of 0.41 on the test set and outperforming PanEcho, a model trained on over one million echocardiographic exams. Source code is available at https://github.com/EIDOSLAB/cardiac-output.