Predictive Diagnostic Approach to Dementia and Dementia Subtypes Using Wireless and Mobile Electroencephalography: A...

Area of Science:

• Neuroscience
• Medical Technology

Background:

• Early detection of mild cognitive impairment and dementia is challenging.
• Electroencephalography (EEG) is a noninvasive brain activity assessment tool.
• Wireless and mobile EEG (wmEEG) offers a tolerable screening method for all dementia stages.

Purpose of the Study:

• To investigate the utility of wmEEG in differentiating dementia from non-dementia controls (NCs).
• To identify specific tasks and brainwave patterns indicative of dementia.
• To assess wmEEG's capability in distinguishing between Alzheimer's disease (AD) and dementia with Lewy bodies (DLB).

Main Methods:

• Recorded bioelectrical activity (BA) using wmEEG from 101 participants (dementia and NCs) during four tasks.
• Performed time-frequency analysis on recorded BA.
• Utilized deep learning for classification analysis to discriminate between dementia and NC, and between AD and DLB.

Main Results:

• Significant differences in power spectra (3-4, 5-7, and 17-23 Hz) were found between dementia and NC under eyes-open conditions.
• A distinct slow alpha power spectrum difference (6-8 Hz) was observed between AD and DLB under eyes-closed conditions.
• Deep learning classification achieved higher accuracy for dementia vs. NC discrimination (0.71) with eyes open and AD vs. DLB discrimination (0.77) with eyes closed.

Conclusions:

• wmEEG is a valuable tool for recording brain activity.
• Analyzing BA with wmEEG can aid in early dementia detection.
• wmEEG facilitates objective and effective discrimination of dementia subtypes.