Congenital Heart Septum Defect Diagnosis on Chest X-Ray Features Using Neural Networks

Abstract

Artificial Neural Network is an information processing paradigm that is inspired by the biological nervous system. Decision Support System (DSS) has been identified as one of the important solution providers in the emerging field of Artificial Neural Networks. Medical Decision Support System (MDSS) is an interactive Decision Support System software, which is designed to assist physicians and other health professionals in decision making tasks and to diagnose the patient disease. The Medical Decision Support System reduces the diagnosis time and improves the accuracy of the diagnosis. One of the clinical tests performed to diagnose Congenital Heart Septum Defect is the Chest Radiography (X-Ray) through the contour of size, position and shape of the heart. In order to diagnose Congenital Heart Septum Defect, a physician analyzes the chest X-ray and extracts the features like heart size measurements. But manual extraction of features and diagnosis is a difficult task for a physician. Therefore, in the present study, an algorithm is developed to automatically analyze and to extract the features from Chest X-ray using Image Processing Techniques. Also, a Decision Support System is developed to Diagnose the Congenital Heart Septum Defect based on chest X-ray features using Backpropagation Neural Network model. The Network is trained by using a Delta Learning Rule. The proposed feature extraction algorithm and Decision Support System are implemented in MATLAB with GUI features.