Processing of echocardiographic images using segmentation, feature extraction and classification for detection of heart abnormality

Abstract

This article is mainly focused to accurately detecting any abnormality of heart if present using echocardiographic image of the patient. Heart abnormalities are now a days very common not only in India but all over the globe irrespective of age and gender. The detection of abnormality is achieved by using Artificial neural network (ANN) Classifier. However, processing of the image is achieved through preprocessing, segmentation, feature extraction and then achieving classification. Processing of image for removal of noise and enhancement is carried out as Preprocessing of image followed by segmentation. The most significant processing task is segmentation which is discussed in detail and preferable algorithm which overcomes the drawbacks and limitations of previous algorithms is proposed. This algorithm is a solution to all problems faced in previous algorithms. carried out using different techniques, three different segmentation techniques are discussed where algorithm proposed Reaction Diffusion Level Set Segmentation (RDLSS) is better than other three methods also overcome the problems faced in previous algorithms, then feature extraction is done to extract energy features where the novelty of the research is use of symlet, Debauches and Bio orthogonal filters for feature extraction and these features are used to classify the images as normal or abnormal using ANN classifier. The ANN classifier is effective and efficient resulting in accuracies of greater than 98%. The results are also clinically validated by doctors.