Alzheimer's Disease Diagnosis and Severity Level Detection Based on Electroencephalography Modulation Spectral...

Area of Science:

• Neuroscience
• Biomedical Engineering
• Medical Diagnostics

Background:

• Electroencephalography (EEG) is a key tool for diagnosing cortical disorders like Alzheimer's disease (AD).
• Traditional analysis of resting-state EEG (rsEEG) relies on conventional frequency bands, but non-conventional bands may offer improved diagnostic accuracy.
• Neuromodulatory deficits are characteristic of AD, necessitating advanced diagnostic markers.

Purpose of the Study:

• To introduce novel features derived from the 2D modulation spectral domain of rsEEG signals for AD characterization.
• To evaluate the efficacy of these new features in discriminating between healthy controls and AD patients across different severity levels.
• To compare the diagnostic performance of the proposed features against traditional rsEEG features.

Main Methods:

• rsEEG signals were analyzed in the 2D modulation spectral domain.
• New features were computed as power within specific regions of interest in the power modulation spectrogram.
• Source localization was employed to validate the biological origin of the proposed features.

Main Results:

• The proposed features demonstrated high discriminant ability for AD severity levels.
• These novel features significantly improved the discrimination between healthy elderly controls and AD patients.
• The features also proved effective in monitoring AD severity, distinguishing between mild and moderate AD.
• Performance of the proposed features surpassed that of traditional rsEEG features.

Conclusions:

• The novel 2D modulation spectral features offer a more effective approach for AD diagnosis and severity monitoring using rsEEG.
• These features provide a biologically validated method for assessing neuromodulatory deficits in AD.
• The findings suggest a potential advancement in non-invasive AD diagnostics through enhanced EEG signal analysis.