Topological Network Analysis of Electroencephalographic Power Maps

Area of Science:

• Neuroscience
• Data Science
• Medical Imaging

Background:

• Meditation offers health benefits, prompting neuroscientific research into its effects on brain activity.
• Electroencephalogram (EEG) is a common tool for studying meditation's neuroplastic effects due to its cost-effectiveness and temporal resolution.
• Standard analysis of EEG spectral power maps often uses statistical methods, potentially missing complex topological information.

Purpose of the Study:

• To introduce a novel inference procedure using topological data analysis (TDA) for EEG signal processing.
• To apply persistent homology to analyze topographic power maps derived from high-density EEG signals.
• To compare the topological features of EEG spectral powers between long-term meditators and meditation-naive individuals.

Main Methods:

• Utilized topological data analysis (TDA), specifically persistent homology, on EEG topographic power maps.
• Developed a novel inference procedure based on sublevel set filtrations of power maps.
• Applied the TDA pipeline to both simulated and real high-density EEG data.

Main Results:

• The study successfully applied the novel TDA pipeline to analyze EEG data.
• Distinct persistent homological features were observed in the high-frequency bands of EEG signals between meditators and non-meditators.
• The findings suggest TDA can offer unique insights into meditation-induced brain activity changes.

Conclusions:

• Topological data analysis, particularly persistent homology, provides a novel approach to understanding EEG spectral power variations.
• The proposed method can differentiate between long-term meditators and meditation-naive practitioners based on their EEG topographic features.
• This research highlights the potential of TDA in advancing neuroscientific investigations of meditation.