Using Neural and Fuzzy Software for the Classification of ECG Signals

Saad Alshaban ； Rawaa Ali

《Research Journal of Applied Sciences, Engineering and Technology》 2卷1期 (2010/01) Pp. 5-10

並列摘要

Two approaches to classify the ECG biomedical signals are presented in this work. One is the Artificial Neural Network (ANN) with multilayer perceptron and the other is the Fuzzy Logic with Fuzzy Knowledge Base Controller (FKBC). Backpropagation Learning Algorithm (BPA) has been used at preset to train the ANN. MATLAB version 6.5 program was used. The ECG signals were classified to eleven groups, one of them is for the normal cases and the others represent ten different diseases. These ECG records were taken for the patients of the Surgical Specialization Center. These ECG records were divided into two groups one for training the systems and the other is for testing them. The performance of both systems, i.e. the ANN and the FL, was evaluated for different examples and Both programs give classification for all the cases. With average percentage of error between the training data group and the testing one is 4.793%. FL system takes fewer time to classify the ECG signals than the ANN because the Knowledge in the NN is automatically acquired by the BPA, but the learning process is relatively slow and the analysis of the trained network was found difficult.

並列關鍵字

Back propagation algorithm BPA ； ECG Signal ； fuzzy software and neural network