Cardiology & Vascular Research

A Medical Imaging Approach for Recognising Mitral Regurgitation Through Machine Learning Methods in Cardiac Imaging

Abstract

Vijay Singh, Anwar Ali Sathio, Saeed Anwar, Raja Vavekanand and R Danish

Mitral regurgitation (MR) is a serious heart valve disease that can have devastating consequences if left untreated. Timely diagnosis and treatment are crucial to prevent further complications, but traditional diagnostic methods pose significant challenges. These methods are not only expensive but also labor-intensive, requiring specialized clinical expertise, which creates barriers to effective MR screening. To address these challenges, we propose a novel semi supervised model for MR classification called CUSSP. CUSSP is designed to process cardiac imaging slices from the 4- chamber view of the heart, utilizing standard computer vision techniques and contrastive models to learn from large amounts of unlabeled data. This approach enables the model to leverage the vast amounts of available imaging data, even when labeled data is scarce. By employing specialized classifiers, CUSSP creates the first automated MR classification system, revolutionizing the diagnosis and treatment of this critical heart condition. The significance of CUSSP lies in its ability to overcome the limitations of traditional diagnostic methods. Cardiac imaging is a complex and time-consuming process, requiring specialized expertise to interpret and diagnose MR. CUSSP automates this process, enabling healthcare professionals to focus on high- value tasks while ensuring accurate and timely diagnoses. The performance of CUSSP is impressive, achieving an F1 score of 0.69 and a ROC-AUC score of 0.88 on a test set of 179 labeled sequences, comprising 154 non-MR and 25 MR cases. These results establish the initial benchmark for this new task, demonstrating the potential of CUSSP to transform MR diagnosis. The CUSSP model is trained using a semi supervised approach, combining the strengths of both supervised and unsupervised learning. This approach enables the model to learn from large amounts of unlabeled data, leveraging the inherent patterns and relationships within the imaging data. By incorporating specialized classifiers, CUSSP achieves high accuracy and robustness, even in the presence of limited labeled data. The use of contrastive models in CUSSP is a key innovation, enabling the model to learn from unlabeled data and adapt to new patterns and variations. This approach allows CUSSP to generalize well to unseen data, ensuring accurate diagnoses even in cases with complex or rare presentations. The implications of CUSSP are far-reaching, with the potential to improve patient outcomes, reduce healthcare costs, and enhance the efficiency of cardiac care. By automating MR diagnosis, CUSSP can help address the growing demand for cardiac imaging services, enabling healthcare providers to focus on high-value tasks and improving patient care. CUSSP represents a significant breakthrough in MR diagnosis, offering a novel semisupervised approach to automated classification. By leveraging computer vision techniques, contrastive models, and specialized classifiers, CUSSP achieves high accuracy and robustness, establishing a new benchmark for MR diagnosis. With its potential to transform cardiac care, CUSSP is poised to make a meaningful impact on patient outcomes and healthcare efficiency.

PDF