Computational intelligence and wavelet transform in walk symmetry analysis using accelerometers

Abstract

Computational intelligence and digital signal processing methods are fundamental mathematical tools widely used in biomedical and engineering applications. Gait symmetry analysis plays a very important role in detecting motion disorders in neurology, rehabilitation, and sports performance. This study focuses on data acquisition using a set of accelerometric sensors to record motion dynamics, ensure time synchronization of signals, and extract features for recognizing distinct motion patterns. The proposed methodology incorporates spectral analysis and digital filtering techniques to eliminate noise and irrelevant frequency components. Motion symmetry analysis is performed using energy distribution, calculated by discrete Fourier and wavelet transforms, with a focus on detailed coefficients at a specified decomposition level. Symmetry estimation is achieved by analyzing the ratio of energy within wavelet bands corresponding to the left and right sides of the body. Furthermore, spatial pattern distribution is analyzed to identify motion asymmetry, with artificial intelligence techniques employed for its evaluation. These results demonstrate the potential of computational intel-ligence in clinical diagnostics of specific neurological disorders.