Routine magnetoencephalography in memory clinic patients: A machine learning approach
Alida A Gouw1,2, Arjan Hillebrand2, Deborah N Schoonhoven1,2
1Alzheimer Center and Department of Neurology, VU University medical center, Amsterdam UMC Amsterdam The Netherlands.
Alzheimer'S & Dementia (Amsterdam, Netherlands)
|September 27, 2021
Summary
Magnetoencephalography (MEG) effectively distinguishes Alzheimer's disease (AD) dementia patients from those with subjective cognitive decline. This brain imaging technique shows significant diagnostic value in memory clinic settings.
Area of Science:
- Neuroscience
- Medical Imaging
- Neurology
Background:
- Routine application of magnetoencephalography (MEG) in memory clinics.
- Assessing the diagnostic value of MEG for differentiating Alzheimer's disease (AD) dementia.
Purpose of the Study:
- To evaluate the utility of MEG in distinguishing AD dementia patients from individuals with subjective cognitive decline (SCD).
- To determine the diagnostic accuracy of MEG spectral measures in a clinical memory setting.
Main Methods:
- MEG recordings were performed on 366 memory clinic patients.
- Source-reconstructed MEG data were analyzed using random forest models.
- Diagnostic accuracy was assessed for discriminating AD dementia (n=40) from SCD (n=40) patients.
Main Results:
- A combination of relative theta and delta power achieved high diagnostic accuracy (0.846).
- Sensitivity and specificity for AD dementia discrimination were 0.855 and 0.837, respectively.
- Hippocampal, thalamic, and temporal-occipital regions were key contributors to the diagnostic model.
Conclusions:
- MEG has been successfully integrated into the diagnostic workup of memory clinic patients.
- MEG provides valuable insights for clinical decision-making in dementia assessment.
