EEG Patterns in Mild Cognitive Impairment (MCI) Patients

Area of Science:

• Neuroscience
• Biomedical Engineering
• Computational Biology

Background:

• Alzheimer's disease (AD) treatment is increasingly focused on early intervention.
• Mild Cognitive Impairment (MCI) is a transitional stage, but not all MCI patients develop AD.
• Distinguishing between MCI patients who will progress to AD and those who will not is a clinical challenge.

Purpose of the Study:

• To apply signal processing and pattern recognition to EEG data for classifying AD patients versus controls.
• To categorize MCI patients into subgroups based on EEG Beta power profiles.
• To predict which MCI patients are most likely to progress to AD.

Main Methods:

• Utilized computer-based signal processing and pattern recognition techniques.
• Analyzed electroencephalogram (EEG) data, specifically focusing on Beta power profiles.
• Developed a classification algorithm for AD, controls, and MCI subgroups.

Main Results:

• Achieved >80% accuracy in classifying AD patients versus controls using EEG.
• Identified two distinct MCI subgroups based on EEG Beta power, one resembling AD patients.
• The classification algorithm correctly predicted the clinical status of 4 out of 6 MCI patients at 2-year follow-up.

Conclusions:

• Automated pattern recognition applied to EEG shows promise for classifying AD.
• EEG analysis can potentially identify MCI patients at higher risk of progressing to AD.
• This approach may serve as a valuable clinical tool for early AD diagnosis and patient stratification.