Neural network based ECG arrhythmia performance optimization

Abstract

Any disturbance in the activity of heart that can cause irregular heart rhythm is known as cardiac arrhythmia. Electrocardiogram (ECG) is one of the most promising tools for detecting different types of arrhythmia, which is necessary until it goes fatal and causes loss of life. For detecting ECG arrhythmia Different techniques have been used till now. But one problem that most of the detection methods have faced is the trade-off between speed and accuracy. So in this thesis work we have tried to create a bridge among these two. We have tried to maintain a good processing speed to detect the ECG abnormality by maintaining a good accuracy rate. In this thesis work we have mainly chosen Neural Network based algorithm to detect arrhythmia but before choosing that we have also worked with two other techniques which are Wavelet packet transform and Non-linear based model. In the detection of the abnormality of ECG beat Leverberg-Marquardt algorithm in the Neural Network model has been used as it is the mostly appraised algorithm for its high processing speed. From this thesis work, the final result creates a balance between the speed and accuracy in the detection of ECG arrhythmia which can be improved in future through more research. So this Neural Network based thesis work comes with an option of choosing a method which balances the detection speed and accuracy as well as comes with a great scope of improvement