A comparison of resting state EEG and structural MRI for classifying Alzheimer's disease and mild cognitive impairment

F. R. Farina, D. D. Emek-Savaş, L. Rueda-Delgado, R. Boyle, H. Kiiski, G. Yener, R. Whelan*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

56 Scopus citations

Abstract

Alzheimer's disease (AD) is the leading cause of dementia, accounting for 70% of cases worldwide. By 2050, dementia prevalence will have tripled, with most new cases occurring in low- and middle-income countries. Mild cognitive impairment (MCI) is a stage between healthy aging and dementia, marked by cognitive deficits that do not impair daily living. People with MCI are at increased risk of dementia, with an average progression rate of 39% within 5 years. There is urgent need for low-cost, accessible and objective methods to facilitate early dementia detection. Electroencephalography (EEG) has potential to address this need due to its low cost and portability. Here, we collected resting state EEG, structural MRI (sMRI) and rich neuropsychological data from older adults (55+ years) with AD, amnestic MCI (aMCI) and healthy controls (~60 per group). We evaluated a range of candidate EEG markers (i.e., frequency band power and functional connectivity) for AD and aMCI classification and compared their performance with sMRI. We also tested a combined EEG and cognitive classification model (using Mini-Mental State Examination; MMSE). sMRI outperformed resting state EEG at classifying AD (AUCs ​= ​1.00 vs 0.76, respectively). However, both EEG and sMRI were only moderately good at distinguishing aMCI from healthy aging (AUCs ​= ​0.67–0.73), and neither method achieved sensitivity above 70%. The addition of EEG to MMSE scores had no added benefit relative to MMSE scores alone. This is the first direct comparison of EEG and sMRI for classification of AD and aMCI.

Original languageEnglish (US)
Article number116795
JournalNeuroimage
Volume215
DOIs
StatePublished - Jul 15 2020

Keywords

  • Alzheimer's disease
  • EEG
  • Machine learning
  • Mild cognitive impairment
  • Resting state
  • Structural MRI

ASJC Scopus subject areas

  • Neurology
  • Cognitive Neuroscience

Fingerprint

Dive into the research topics of 'A comparison of resting state EEG and structural MRI for classifying Alzheimer's disease and mild cognitive impairment'. Together they form a unique fingerprint.

Cite this