Bispectrum-based Cross-frequency Functional Connectivity: Classification of Alzheimer's disease

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Alzheimer's disease (AD) is a neurodegenerative condition impacting brain functional connectivity (FC).
• Current research often analyzes neurophysiological signals, like electroencephalography (EEG), within isolated frequency bands, potentially missing crucial cross-frequency interactions.
• Linear FC measures have limitations in capturing the complex dynamics of brain activity in AD.

Purpose of the Study:

• To introduce and evaluate a novel approach for quantifying cross-frequency functional connectivity (FC) in Alzheimer's disease.
• To compare the diagnostic performance of nonlinear cross-frequency FC measures with traditional linear, within-frequency FC measures.
• To establish the clinical relevance of cross-frequency coupling for AD diagnosis.

Main Methods:

• Utilized cross-bispectrum, a higher-order spectral analysis technique, to measure nonlinear cross-frequency FC.
• Compared cross-bispectrum (nonlinear) with cross-spectrum (linear) measures of FC within and across frequency bands.
• Constructed FC networks from frequency coupling data, vectorized them, and used them to train a machine learning classifier.
• Investigated the impact of fusing features from different FC networks on classification accuracy.

Main Results:

• Both within-frequency and cross-frequency FC networks demonstrated high accuracy in predicting Alzheimer's disease.
• FC networks derived from cross-bispectrum analysis significantly outperformed those based on cross-spectrum analysis.
• Fusing features from multiple FC networks further enhanced classification accuracy.
• Cross-frequency FC, particularly nonlinear coupling measured by bispectrum, plays a crucial role in AD.

Conclusions:

• Cross-frequency functional connectivity analysis, especially nonlinear measures like bispectrum, offers a more sensitive approach for detecting Alzheimer's disease.
• The findings highlight the diagnostic relevance of cross-frequency coupling in neurodegenerative diseases like AD.
• This advanced FC analysis method holds promise for improved diagnostic tools in Alzheimer's disease research.