Left ventricle detection in echocardiography videos

Abstract

In this paper, we propose a new fully automated approach of Left Ventricle (LV) segmentation in 2D echocardiography videos. The proposed segmentation method combines texture gradients in moving sequences. The novelty of our approach consists in exploiting simultaneously the intra-frame and the motion information. Experiments are carried out on apical echocardiographic sequences and indicate the benefit that can be drawn from the proposed method in terms of both accuracy and computational complexity.