Analysis of EEG Data Using Complex Geometric Structurization.

Area of Science:

• Neuroscience
• Dynamical Systems
• Computational Geometry

Background:

• Electroencephalogram (EEG) records brain electrical activity, often analyzed as time series data.
• Epilepsy EEG data may exhibit chaotic dynamics, posing interpretation challenges.
• Dynamical systems theory, specifically Embedding Theory, offers methods to reconstruct phase space from time series.

Purpose of the Study:

• To propose a novel approach, complex geometric structurization, for analyzing epilepsy EEG time series.
• To reconstruct phase space representations of brain activity using Embedding Theory.
• To investigate the utility of complex geometric structures as biomarkers for epilepsy seizure changes.

Main Methods:

• Utilized Embedding Theory from dynamical systems to reconstruct phase space from EEG time series.
• Applied complex geometric structurization, employing α-shapes from shape analysis.
• Constructed strange attractors to represent the geometric structures of brain activity.

Main Results:

• Demonstrated proof of concept by showing complex structures capture expected brain lobe activity changes.
• Initial analyses indicate the potential of these structures to reflect dynamic brain changes.
• The method successfully generated complex geometric structures from EEG data.

Conclusions:

• Complex geometric structurization offers a novel framework for analyzing complex time series data like EEG.
• The derived complex geometric structures show promise as potential biomarkers for detecting seizure-related brain changes.
• This approach could enhance the understanding and diagnosis of neurological conditions like epilepsy.