Abnormality in Mitral Leaflets Mobility Detection Using Deep Learning

Main Article Content

Vishal Chandra, Prattay Guha Sarkar, Vinay Singh


RHD (Rheumatic heart disease) is still a burden in underdeveloped countries. It affects the heart's wall and valves, mainly the mitral valve. TTE (Transthoracic Echocardiography) is a non-invasive technique used to look inside the heart, including the heart valve and valve's motion. RHD leads Mitral stenosis. Guidelines for assessing the mitral valve based upon the mitral valve's morphology include leaflets' mobility, leaflets' thickness, and sub valvular thickness. Sonographers able to track the mitral leaflets' motion using the M-mode of mitral valve (MV) in the long-axis view. In this research, we used an M-mode echocardiography view for the mitral valve's leaflet's motion. This research used the deep learning technique for detection abnormality in the mitral leaflets' movement automatically. Mitral stenosis restricts the mitral leaflets’ motion. Our proposed deep learning (DL) model achieved the testing accuracy of 96% having f1 score of 96% on the testing dataset of 64 patients. Our training dataset is of 320 people. Even smaller dataset proposed model gives a promising result. In a normal mitral valve, leaflets flip two times during diastole (mid diastole). else ways, in the Rheumatic mitral valve, there will minimal forward movement. So, there is a distinguishable pattern between normal, and RHD affected mitral valve can see.

Article Details

How to Cite
Vishal Chandra, Prattay Guha Sarkar, Vinay Singh. (2021). Abnormality in Mitral Leaflets Mobility Detection Using Deep Learning. Annals of the Romanian Society for Cell Biology, 5956–5963. Retrieved from https://www.annalsofrscb.ro/index.php/journal/article/view/759