Novel PCG Analysis Method for Discriminating Between Abnormal and Normal Heart Sounds - 28/12/19
Cet article a été publié dans un numéro de la revue, cliquez ici pour y accéder
Graphical abstract |
Highlights |
• | Collection of heart sound database and Features extraction from each PCG signal. |
• | Performing two types of analysis with and without normalization. |
• | Data compression using principal component analysis (PCA). |
• | Classification using SVM and k-NN. |
• | The best obtained accuracies were between 98% and 100% |
Abstract |
A novel approach for separation among normal and heart murmurs sounds based on Phonocardiogram (PCG) analysis is introduced in this paper. The purpose of this work is to find the appropriate algorithm able to detect heart failures. Different features have been extracted from time and frequency domains. After the normalization step, the Principal Component Analysis algorithm is used for data reduction and compression. Support Vectors Machine (SVM), and k-Nearest Neighbors (kNN) classifiers were used with different kernels and number of neighbors in the classification step. Simulation results obtained from different databases are compared. The developed system gave good results when applied to different datasets. The accuracy of 96%, and 100% for the first, and the second dataset respectively were obtained. The algorithm shows its effectiveness in separation between normal and pathological cases.
Le texte complet de cet article est disponible en PDF.Keywords : Phonocardiogram, Signal processing, MFCC, SVM, k-NN, Classification
Plan
Bienvenue sur EM-consulte, la référence des professionnels de santé.
L’accès au texte intégral de cet article nécessite un abonnement.
Déjà abonné à cette revue ?