Stability of MEG and EEG Spectral Features in Mild Cognitive Impairment (MCI)
Poster No:
162
Submission Type:
Abstract Submission
Authors:
Shrikanth Kulashekhar1,2, Antti Kinnunen1, Timo Saarinen1, Ville Mäntynen1, Jaakko Hotta3, Anne Koivisto3, Mia Liljeström1,2, Hanna Renvall1,2
Institutions:
1BioMag Laboratory, HUS Diagnostic Center, Helsinki University Hospital (HUS), Helsinki, Finland, 2Department of Neuroscience and Biomedical Engineering, Aalto University, Helsinki, Finland, 3Department of Clinical Neurosciences, Neurology, Helsinki University Hospital (HUS), Helsinki, Helsinki
First Author:
Shrikanth Kulashekhar
BioMag Laboratory, HUS Diagnostic Center, Helsinki University Hospital (HUS)|Department of Neuroscience and Biomedical Engineering, Aalto University
Helsinki, Finland|Helsinki, Finland
BioMag Laboratory, HUS Diagnostic Center, Helsinki University Hospital (HUS)|Department of Neuroscience and Biomedical Engineering, Aalto University
Helsinki, Finland|Helsinki, Finland
Co-Author(s):
Antti Kinnunen
BioMag Laboratory, HUS Diagnostic Center, Helsinki University Hospital (HUS)
Helsinki, Finland
BioMag Laboratory, HUS Diagnostic Center, Helsinki University Hospital (HUS)
Helsinki, Finland
Timo Saarinen
BioMag Laboratory, HUS Diagnostic Center, Helsinki University Hospital (HUS)
Helsinki, Finland
BioMag Laboratory, HUS Diagnostic Center, Helsinki University Hospital (HUS)
Helsinki, Finland
Ville Mäntynen
BioMag Laboratory, HUS Diagnostic Center, Helsinki University Hospital (HUS)
Helsinki, Finland
BioMag Laboratory, HUS Diagnostic Center, Helsinki University Hospital (HUS)
Helsinki, Finland
Jaakko Hotta
Department of Clinical Neurosciences, Neurology, Helsinki University Hospital (HUS)
Helsinki, Helsinki
Department of Clinical Neurosciences, Neurology, Helsinki University Hospital (HUS)
Helsinki, Helsinki
Anne Koivisto
Department of Clinical Neurosciences, Neurology, Helsinki University Hospital (HUS)
Helsinki, Helsinki
Department of Clinical Neurosciences, Neurology, Helsinki University Hospital (HUS)
Helsinki, Helsinki
Mia Liljeström
BioMag Laboratory, HUS Diagnostic Center, Helsinki University Hospital (HUS)|Department of Neuroscience and Biomedical Engineering, Aalto University
Helsinki, Finland|Helsinki, Finland
BioMag Laboratory, HUS Diagnostic Center, Helsinki University Hospital (HUS)|Department of Neuroscience and Biomedical Engineering, Aalto University
Helsinki, Finland|Helsinki, Finland
Hanna Renvall
BioMag Laboratory, HUS Diagnostic Center, Helsinki University Hospital (HUS)|Department of Neuroscience and Biomedical Engineering, Aalto University
Helsinki, Finland|Helsinki, Finland
BioMag Laboratory, HUS Diagnostic Center, Helsinki University Hospital (HUS)|Department of Neuroscience and Biomedical Engineering, Aalto University
Helsinki, Finland|Helsinki, Finland
Introduction:
In mild cognitive impairment (MCI), the first neurodegenerative changes may occur years be-fore they can be effectively detected with current clinical tools. Early pathology has been linked to synaptic dysfunction causing brain network disturbances that may be observed with electroencephalography (EEG) and magnetoencephalography (MEG) (Pusil S, 2019; Miraglia F, 2020). However, current clinical practice lacks critical tools for early identification of those at greatest risk of developing into clinical dementia, and in need of preventive actions.
In a 5-year EU Horizon2020 project 'AI-Mind', we collect EEG and MEG data from 1000 participants with MCI in Europe (Finland, Italy, Norway, and Spain), to be compiled with cognitive, genetic and plasma biomarker measures and techniques based on artificial intelligence (AI) for predicting the overall dementia risk. We aim at earlier and more accurate interception of those MCI individuals in a 'prodromal' stage of dementia, to allow timely intervention for known modifiable risk factors (Figure 1).
Various features of the MEG and EEG signals have been compared between MCI patients and healthy controls, including the peak frequency, peak and average power, at different frequency bands. The occipital alpha (8-13 Hz) rhythm has been shown to reduce in power, slow down in its peak frequency, and change in its cortical distribution in MCI (Kowalski JW, 2001). However, the stability and reliability of these measures have not been addressed in large subject cohorts nor in MCI. A good biomarker should be stable within a measurement in all subjects, and, in case of MCI, show differences between measurements separated in time, but only in subjects who are in risk of disease progression.
In a 5-year EU Horizon2020 project 'AI-Mind', we collect EEG and MEG data from 1000 participants with MCI in Europe (Finland, Italy, Norway, and Spain), to be compiled with cognitive, genetic and plasma biomarker measures and techniques based on artificial intelligence (AI) for predicting the overall dementia risk. We aim at earlier and more accurate interception of those MCI individuals in a 'prodromal' stage of dementia, to allow timely intervention for known modifiable risk factors (Figure 1).
Various features of the MEG and EEG signals have been compared between MCI patients and healthy controls, including the peak frequency, peak and average power, at different frequency bands. The occipital alpha (8-13 Hz) rhythm has been shown to reduce in power, slow down in its peak frequency, and change in its cortical distribution in MCI (Kowalski JW, 2001). However, the stability and reliability of these measures have not been addressed in large subject cohorts nor in MCI. A good biomarker should be stable within a measurement in all subjects, and, in case of MCI, show differences between measurements separated in time, but only in subjects who are in risk of disease progression.
·Figure 1:The AI-Mind pipeline.
Methods:
Here, data from 41 Finnish MCI subjects (20 females, age 68 ± 5; mean ± SD) were used. These subjects were recruited via HUS neurology departments and general advertisements to participate in the AI-Mind study and were initially screened by a neurologist in reference to the inclusion and exclusion criteria of the study. The subjects underwent a combined MEG and EEG resting-state measurement consisting of two sessions of 5 min of eyes open and two sessions of 5 min with their eyes closed.
MEG recordings were conducted at the BioMag Laboratory in Helsinki University Hospital (HUS) with a 306-channel neuromagnetometer (Triux, MEGIN); the EEG recordings were conducted simultaneously with the MEG with 128-channel system (eegoTM, eemagine Medical Imaging Solutions GmbH).
The data analysis was performed using custom Python software developed in BioMag Laboratory and external software such as Maxfilter, Freesurfer, FOOOF, and MNE python. We determined the within-session test-retest reliability for spectral peak frequencies and relative peak power (Figure 2A). The peak frequency and power were computed for the alpha-band (8-13 Hz) over the occipital brain regions and for the beta-band (14-30 Hz) over the fronto-central brain regions. For each subject the peak frequency and power were automatically identified and visually verified. Finally, the test-retest stability of the spectral features was accessed using intraclass correlation (ICC) method.
MEG recordings were conducted at the BioMag Laboratory in Helsinki University Hospital (HUS) with a 306-channel neuromagnetometer (Triux, MEGIN); the EEG recordings were conducted simultaneously with the MEG with 128-channel system (eegoTM, eemagine Medical Imaging Solutions GmbH).
The data analysis was performed using custom Python software developed in BioMag Laboratory and external software such as Maxfilter, Freesurfer, FOOOF, and MNE python. We determined the within-session test-retest reliability for spectral peak frequencies and relative peak power (Figure 2A). The peak frequency and power were computed for the alpha-band (8-13 Hz) over the occipital brain regions and for the beta-band (14-30 Hz) over the fronto-central brain regions. For each subject the peak frequency and power were automatically identified and visually verified. Finally, the test-retest stability of the spectral features was accessed using intraclass correlation (ICC) method.
Results:
The results indicate good (ICC>0.6) test-retest reliability across runs for EEG (peak frequency: 0.72 (EC); relative power: 0.86 (EC) and 0.66 (EO); p<0.001) and excellent test-retest reliability (ICC>0.75) for MEG (peak frequency: 0.83 (EC); relative power: 0.96 (EC) and 0.85 (EO); p<0.001). Consistency across EEG and MEG was excellent (peak frequency: 0.86 (EC); relative power 0.89 (EC) and 0.84 (EO); p<0.001).
·Figure 2:(A) Features in MEG and EEG. (B) Stability of features between run 1 and run 2.
Conclusions:
Electrophysiological recordings of brain networks are emerging as a diagnostic tool for early detection of dementia. Good stability of the measured signal features is a prerequisite for diagnosis. Here, we demonstrate good or excellent stability for prominent spectral features both within and across EEG and MEG recordings.
Disorders of the Nervous System:
Neurodegenerative/ Late Life (eg. Parkinson’s, Alzheimer’s) 1
Modeling and Analysis Methods:
EEG/MEG Modeling and Analysis 2
Task-Independent and Resting-State Analysis
Novel Imaging Acquisition Methods:
EEG
MEG
Keywords:
Aging
Data analysis
Degenerative Disease
Electroencephaolography (EEG)
MEG
Source Localization
1|2Indicates the priority used for review
Provide references using author date format
Miraglia F. (2020), ‘Small World Index in Default Mode Network Predicts Progression from Mild Cognitive Impairment to Dementia’, International Journal of Neural Systems, 30: 2050004.
Pusil S. (2019), ‘Aberrant MEG multi-frequency phase temporal synchronization predicts conversion from mild cognitive impairment-to-Alzheimer's disease’, Neuroimage Clin 24:101972.