Interpretation of ECG using Modified Intuitionistic Fuzzy C-Means Clustering for Arrhythmia Data

An electrocardiogram (ECG) is defined as a measure of variation in the electrical activity of the heart and is broadly used in detection and classification of heart-related diseases. The abnormalities present in the heart can be easily analyzed through the variation in electrical signal captured from the heart through impulse waveforms which are generated by certain specialized cardiac tissues. Different authors have developed various clustering models and classification techniques for detecting heart-related diseases. However there still exists a limitation in terms of accuracy. In this article, we proposed a new modified unsupervised clustering algorithm for effective detection of heart diseases. To select the best discriminate feature for effective learning, this article make use of feature selection methods such as principal component analysis, linear discriminative analysis, and regularized locality preserving indexing. The reduced features set are clustered using modified intuitionistic Fuzzy C-means clustering (mifcm) method. The experiment results proved that the proposed method effectively identifies the discriminative features. Further the obtained accuracy is also better when compared to other existing method.