Analysis of knee joint vibroarthrographic signals using the ensemble empirical mode decomposition and variational mode decomposition

Knee injuries are one of the most common injuries in people s daily lives especially in vibrant sports. Analysis of vibroarthrograph signals has shown promise for the non-invasive diagnosis of knee joint disorders. In this paper compares two approaches for denoising VAG signals by using a new method of variational mode decomposition followed by wiener entropy thresholding and Ensemble Empirical Mode Decomposition. Also, is used six type of features in both time and frequency domain in order to discriminate normal and abnormal signals. A classification accuracy of 86.56%±0.54 and 85.36%±0.23 is obtained by variational mode decomposition and Ensemble Empirical Mode Decomposition with a database of 89 VAG signals by radial basis function neural network

Keywords:

vibroarthrography , Ensemble Empirical Mode Decomposition , variational mode decomposition , radial basis function neural network