BSS and ICA in neuroinformatics: from current practices to open challenges

Area of Science:

• Neuroinformatics
• Computational Neuroscience
• Signal Processing

Background:

• Neuroimaging techniques like electroencephalography (EEG) and functional magnetic resonance imaging (fMRI) generate complex data.
• Analyzing these signals requires advanced methods to isolate meaningful brain activity from noise and artifacts.

Purpose of the Study:

• To provide an overview of Blind Source Separation (BSS) and Independent Component Analysis (ICA) applications in neuroinformatics.
• To highlight the utility of BSS/ICA in processing electrophysiological recordings and fMRI data.
• To discuss current challenges and future directions in applying these techniques to brain data.

Main Methods:

• Application of BSS and ICA algorithms to electrophysiological recordings (e.g., EEG).
• Application of BSS and ICA algorithms to functional magnetic resonance images (fMRI).
• Illustrative examples including artifact identification/removal and fMRI data analysis.

Main Results:

• Demonstrated successful identification and removal of artifacts in both electrophysiological and fMRI data.
• Showcased the analysis of a simple fMRI dataset using BSS/ICA.
• Identified key open challenges in signal processing for neuroinformatics.

Conclusions:

• BSS and ICA are valuable techniques for neuroinformatics, particularly for artifact removal and signal analysis in EEG and fMRI.
• Future research should focus on independent subspace analysis, functional brain network studies, and single-trial analysis.
• Addressing these challenges will enhance our understanding of brain activity from complex neuroimaging data.