Wavelet transform analysis of vibroarthrographic (VAG) signals obtained during dynamic knee movement

The vibroarthrographic (VAG) signal can be measured on the skin surface of human knee joints during their relative movements. The nonstationary nature of the VAG signal makes frequency domain analysis using conventional Fourier methods difficult. Wavelet transform (WT) analysis is proposed. Results show that the method of multiresolution decomposition can localize events in the joint angle (time) and scale (frequency) planes. This dual localization property of the WT is particularly useful in studying VAG signals, where frequency spectra at various joint angles are required for diagnostic purposes