ECG Data Mining Approach for Detection of Arrhythmia Using Machine Learning

The Electrocardiogram (ECG) is very important signal generated from human body that provides information for assessment of variation in cardiac activities. These ECG data are generally distorted due to high power line or magnetic field associated noise that results as a problem in error during diagnosis of heart related disorders on analysis through biomedical analysis software. This work contributes for very significant stage that is introduced to reduce the anomaly in the ECG signals to perform accurate automated diagnosis. This work follows a methodology of an automated machine learning algorithm is developed for accurate estimation of ECG data segment cycle using clustering approach integrated with Discrete Wavelet Transform (DWT). The key findings of this works cover the development and analysis of an approach to overcome the errors in detection of T-wave using Implantable Cardioverter Defibrillator (ICD). The work reports the solution of issues using fuzzy C-means clustering behind prediction of fatal Ventricular Arrhythmias (VA) for performing standard classifications.