Computational methods of EEG signals analysis for Alzheimer's disease classification

Area of Science:

• Neuroscience
• Computational Biology
• Medical Diagnostics

Background:

• Alzheimer's disease (AD) is a progressive neurological disorder causing cognitive decline.
• Early diagnosis of AD is crucial for patient quality of life, though no cure exists.
• Electroencephalographic (EEG) signal analysis offers a potential avenue for non-invasive diagnostic tools.

Purpose of the Study:

• To evaluate the efficacy of six computational time-series analysis methods for detecting Alzheimer's disease using EEG data.
• To identify which specific methods demonstrate robust discrimination between AD patients and healthy controls.

Main Methods:

• Applied six computational time-series analysis methods: wavelet coherence, fractal dimension, quadratic entropy, wavelet energy, quantile graphs, and visibility graphs.
• Utilized EEG records from 160 AD patients and 24 healthy elderly controls.
• Analyzed both raw and wavelet-filtered (alpha, beta, theta, delta bands) EEG signals.

Main Results:

• Wavelet coherence and quantile graphs demonstrated robust ability to differentiate AD patients from healthy elderly subjects.
• The applied computational methods showed potential in analyzing EEG signals for AD detection.
• Analysis of filtered EEG bands (alpha, beta, theta, delta) provided insights into AD-related signal changes.

Conclusions:

• Computational time-series analysis of EEG signals, particularly wavelet coherence and quantile graphs, offers a promising non-invasive approach for Alzheimer's disease detection.
• These methods represent a low-cost diagnostic strategy for identifying AD in elderly populations.
• Further research can refine these techniques for clinical application in early AD diagnosis.