Adaptive Data Analysis Methods for Biomedical Signal Processing Applications

Abstract

Biomedical signals represent the variation in electric potential due to physiological processes and are recorded through certain types of sensors or electrodes. In practice, the biomedical signals are typically complex and non-stationary. This makes adaptive data-driven techniques a natural choice for processing biomedical signals. Signal processing methods such as the Fourier transform make use of some pre-defined basic functions designed independent of the signal information. Data-driven methods propose such basic functions directly depending on the information content in the signal. The adaptive data analysis methods tend to decompose a signal into individual modes that are present in it, thus separating them from each other. This chapter presents a detailed review of adaptive data analysis techniques including wavelet transform, empirical mode decomposition, empirical wavelet transform, and variational mode decomposition with their applications to biomedical signal analysis.