MFCC Feature Extraction and KNN Classification in ECG Signals

Feature extraction of electrocardiogram (ECG) signal is one of the essential steps to diagnose various cardiovascular disease (CVD). The signal is generated by the hearts electrical activity and able to reveal the abnormal activity of the heart. An accurate feature extraction method is important to produce better identification of ECG signal. ECG feature extraction using Mel Frequency Cepstrum Coefficient (MFCC), Discrete Wavelet transformation and KNN using euclidean distance as the classifier is proposed in this study. The model and testing of the proposed system were performed on the two types of data, normal and myocardial infarction (MI) labeled as abnormal, obtained from PTB-DB database. Total data used were 100 data, with 50 data for each condition. K-fold cross validation also applied to achieve a generalized result. According to the experimental, 13 features that obtained from MFCC shows good result. The accuracy, sensitivity and specificity were achieved 84%, 85% and 84% respectively.

[1]  Min Zhou,et al.  ECG Classification Using Wavelet Packet Entropy and Random Forests , 2016, Entropy.

[2]  Shweta H. Jambukia,et al.  Classification of ECG signals using machine learning techniques: A survey , 2015, 2015 International Conference on Advances in Computer Engineering and Applications.

[3]  Emran Mohd Tamil,et al.  Heartbeat Electrocardiogram (ECG) Signal Feature Extraction Using Discrete Wavelet Transforms (DWT) , 2008 .

[4]  Ali Işın,et al.  Cardiac arrhythmia detection using deep learning , 2017 .

[5]  Derya Avci,et al.  A new technique for ECG signal classification genetic algorithm Wavelet Kernel extreme learning machine , 2019, Optik.

[6]  U. Rajendra Acharya,et al.  Application of deep convolutional neural network for automated detection of myocardial infarction using ECG signals , 2017, Inf. Sci..

[7]  Adriana Mexicano,et al.  Feature extraction of electrocardiogram signals by applying adaptive threshold and principal component analysis , 2015 .

[8]  Arun Khosla,et al.  QRS detection using K-Nearest Neighbor algorithm (KNN) and evaluation on standard ECG databases , 2012, Journal of advanced research.

[9]  Mohamed Atibi,et al.  ECG signals classification using MFCC coefficients and ANN classifier , 2016, 2016 International Conference on Electrical and Information Technologies (ICEIT).

[10]  R VijayaArjunan.  ECG SIGNAL CLASSIFICATION BASED ON STATISTICAL FEATURES WITH SVM CLASSIFICATION , 2016 .

[11]  Achmad Rizal.  Perbandingan Skema Dekomposisi Paket Wavelet untuk Pengenalan Sinyal EKG , 2015 .

[12]  Risanuri Hidayat,et al.  Denoising Speech for MFCC Feature Extraction Using Wavelet Transformation in Speech Recognition System , 2018, 2018 10th International Conference on Information Technology and Electrical Engineering (ICITEE).

[13]  Engin Avci,et al.  Intelligent system based on Genetic Algorithm and support vector machine for detection of myocardial infarction from ECG signals , 2018, 2018 26th Signal Processing and Communications Applications Conference (SIU).

[14]  Vijaya R Arjunan,et al.  ECG SIGNAL CLASSIFICATION BASED ON STATISTICAL FEATURES WITH SVM CLASSIFICATION , 2017 .

[15]  Behnam Farzam,et al.  The Diagnosis of Heart Diseases Based on PCG Signals using MFCC Coefficients and SVM Classifier , 2014 .