Multivariate adaptive signal decomposition techniques and their applications to EEG signal processing: An introduction

Abstract

Due to tremendous improvements in sensing technology, simultaneously acquiring signals from multiple locations becomes possible. Recording signals using multiple electrodes improves the spatial resolution of the signal. The availability of computational resources allows us to process these signals efficiently. But processing these multivariate signals in a univariate way results in a loss of mutual information present in the signal. Univariate signal decomposition techniques have been extended to multivariate signals to address these issues. The multivariate extensions are developed in such a manner that they can generate the same number of multivariate oscillatory modes across different channels and have similar frequency components. In this chapter, we have described the multivariate extension of adaptive signal decomposition algorithms with the help of steps. The decomposed oscillatory components for the synthetic signal and the real-time electroencephalogram (EEG) signal are shown for each algorithm. Finally, the applications of these multivariate adaptive signal decompositions are discussed. © 2025 Elsevier Inc. All rights are reserved including those for text and data mining AI training and similar technologies.