Machine Learning for Automated Mitral Regurgitation Detection from Cardiac Imaging

TL;DR

This paper introduces CUSSP, a semi-supervised machine learning model that automates mitral regurgitation detection from cardiac images, reducing reliance on expert labeling and achieving promising diagnostic accuracy.

Contribution

The study presents the first automated MR classification system using semi-supervised learning on cardiac imaging, leveraging contrastive models and standard computer vision techniques.

Findings

Achieved an F1 score of 0.69 on test data.

Attained a ROC-AUC score of 0.88.

Established the first benchmark for automated MR detection.

Abstract

Mitral regurgitation (MR) is a heart valve disease with potentially fatal consequences that can only be forestalled through timely diagnosis and treatment. Traditional diagnosis methods are expensive, labor-intensive and require clinical expertise, posing a barrier to screening for MR. To overcome this impediment, we propose a new semi-supervised model for MR classification called CUSSP. CUSSP operates on cardiac imaging slices of the 4-chamber view of the heart. It uses standard computer vision techniques and contrastive models to learn from large amounts of unlabeled data, in conjunction with specialized classifiers to establish the first ever automated MR classification system. Evaluated on a test set of 179 labeled -- 154 non-MR and 25 MR -- sequences, CUSSP attains an F1 score of 0.69 and a ROC-AUC score of 0.88, setting the first benchmark result for this new task.