Automated Inception Network based Cardiac Image Segmentation Analysis

Abstract

Coronary Heart disease is one of the main threats to human health, and approximately ten million people die from heart disease annually throughout the world. Magnetic resonance imaging is widely used for cardiopathy diagnosis. Left ventricle Segmentation from cardiac magnetic resonance imaging (MRI) is considered as significant process for estimation of clinical indices such as stroke volume, ejection fraction. This research paper proposed a new automatic LV segmentation method, which is attained by using Inception based Convolutional Neural Network (ICNN). The technique uses the inception network power, which is combined with convolutional layers in deep neural network architecture. The research study empirically evaluated the segmentation accuracy of the proposed ICNN model by using publicly available benchmark dataset. The experimental results stated that the developed method achieved 94.23% segmentation accuracy than existing neural networks.