Data-driven retrieval of population-level EEG features and their role in neurodegenerative diseases DOI Creative Commons
Wentao Li, Yogatheesan Varatharajah, Ellen Dicks

et al.

Brain Communications, Journal Year: 2024, Volume and Issue: 6(4)

Published: Jan. 1, 2024

Abstract Electrophysiologic disturbances due to neurodegenerative disorders such as Alzheimer’s disease and Lewy Body are detectable by scalp EEG can serve a functional measure of severity. Traditional quantitative methods analysis often require an a-priori selection clinically meaningful features susceptible bias, limiting the clinical utility routine EEGs in diagnosis management disorders. We present data-driven tensor decomposition approach extract top 6 spectral spatial representing commonly known sources activity during eyes-closed wakefulness. As part their neurologic evaluation at Mayo Clinic, 11 001 patients underwent 12 176 routine, standard 10–20 studies. From these raw EEGs, we developed algorithm based on posterior alpha eye movement automatically select awake-eyes-closed epochs estimated average power density (SPD) between 1 45 Hz for each channel. then created three-dimensional (3D) (record × channel frequency) applied canonical polyadic six factors. further identified independent cohort meeting consensus criteria mild cognitive impairment (30) or dementia (39) with Bodies (31) similarly aged cognitively normal controls (36). evaluated ability factors differentiating subgroups using Naïve Bayes classification assessed linear associations factor loadings Kokmen short test mental status scores, fluorodeoxyglucose (FDG) PET uptake ratios CSF Disease biomarker measures. Factors represented biologically brain activities including rhythm, anterior delta/theta rhythms centroparietal beta, which correlated patient age dysrhythmia grade. These were also able distinguish from moderate high degree accuracy (Area Under Curve (AUC) 0.59–0.91) (AUC 0.61). Furthermore, relevant performance, metabolism AB42 measures subgroup. This study demonstrates that approaches population-level without artefact rejection channels frequency bands. With continued development, may improve memory care assisting early identification different causes impairment.

Language: Английский

Aim-based choice of strategy for MEG-based brain state classification DOI

Irina Saranskaia,

Boris Gutkin, Denis Zakharov

et al.

The European Physical Journal Special Topics, Journal Year: 2025, Volume and Issue: unknown

Published: March 22, 2025

Language: Английский

Citations

0
Near-death experience during cardiac arrest and consciousness beyond the brain: a narrative review DOI

Bruno Angeli-Faez,

Bruce Greyson,

Pim van Lommel

et al.

International Review of Psychiatry, Journal Year: 2025, Volume and Issue: unknown, P. 1 - 12

Published: May 15, 2025

Language: Английский

Citations

0
The impact of ROI extraction method for MEG connectivity estimation: Practical recommendations for the study of resting state data. DOI Creative Commons
Diandra Brkić, Sara Sommariva, Anna‐Lisa Schuler

et al.

NeuroImage, Journal Year: 2023, Volume and Issue: 284, P. 120424 - 120424

Published: Oct. 30, 2023

Magnetoencephalography and electroencephalography (M/EEG) seed-based connectivity analysis requires the extraction of measures from regions interest (ROI). M/EEG ROI-derived source activity can be treated in different ways. It is possible, for instance, to average each ROI's time series prior calculating measures. Alternatively, one compute maps element ROI dimensionality reduction obtain a single map. The impact these strategies on results still unclear. Here, we address this question within large MEG resting state cohort (N=113) simulated data. We consider 68 ROIs (Desikan-Kiliany atlas), two (phase locking value-PLV, its imaginary counterpart- ciPLV), three frequency bands (theta 4-8 Hz, alpha 9-12 beta 15-30 Hz). compare four methods: (i) mean, or (ii) PCA seed before computing connectivity, map (iii) average, (iv) maximum after seed. Hierarchical clustering then applied outputs across multiple strategies, followed by direct contrasts methods. Finally, are validated using set realistic simulations. show that ROI-based vary remarkably terms magnitude spatial distribution. Dimensionality procedures conducted more similar each-other, while approach most dissimilar other approaches. Although differences methods consistent bands, they influenced metric size. Greater were observed ciPLV than PLV, larger ROIs. Realistic simulations confirmed aggregation generally accurate but have lower specificity (higher rate false positive connections). Though computationally demanding, should preferred when higher sensitivity desired. Given remarkable procedures, caution warranted comparing studies applying

Language: Английский

Citations

9
Prefrontal oscillatory slowing in early-course schizophrenia is associated with worse cognitive performance and negative symptoms: a TMS-EEG study DOI
Francesco Donati,

Ahmad Mayeli,

Bruno Andry Nascimento Couto

et al.

Biological Psychiatry Cognitive Neuroscience and Neuroimaging, Journal Year: 2024, Volume and Issue: unknown

Published: July 1, 2024

Language: Английский

Citations

3
Data-driven retrieval of population-level EEG features and their role in neurodegenerative diseases DOI Creative Commons
Wentao Li, Yogatheesan Varatharajah, Ellen Dicks

et al.

Brain Communications, Journal Year: 2024, Volume and Issue: 6(4)

Published: Jan. 1, 2024

Abstract Electrophysiologic disturbances due to neurodegenerative disorders such as Alzheimer’s disease and Lewy Body are detectable by scalp EEG can serve a functional measure of severity. Traditional quantitative methods analysis often require an a-priori selection clinically meaningful features susceptible bias, limiting the clinical utility routine EEGs in diagnosis management disorders. We present data-driven tensor decomposition approach extract top 6 spectral spatial representing commonly known sources activity during eyes-closed wakefulness. As part their neurologic evaluation at Mayo Clinic, 11 001 patients underwent 12 176 routine, standard 10–20 studies. From these raw EEGs, we developed algorithm based on posterior alpha eye movement automatically select awake-eyes-closed epochs estimated average power density (SPD) between 1 45 Hz for each channel. then created three-dimensional (3D) (record × channel frequency) applied canonical polyadic six factors. further identified independent cohort meeting consensus criteria mild cognitive impairment (30) or dementia (39) with Bodies (31) similarly aged cognitively normal controls (36). evaluated ability factors differentiating subgroups using Naïve Bayes classification assessed linear associations factor loadings Kokmen short test mental status scores, fluorodeoxyglucose (FDG) PET uptake ratios CSF Disease biomarker measures. Factors represented biologically brain activities including rhythm, anterior delta/theta rhythms centroparietal beta, which correlated patient age dysrhythmia grade. These were also able distinguish from moderate high degree accuracy (Area Under Curve (AUC) 0.59–0.91) (AUC 0.61). Furthermore, relevant performance, metabolism AB42 measures subgroup. This study demonstrates that approaches population-level without artefact rejection channels frequency bands. With continued development, may improve memory care assisting early identification different causes impairment.

Language: Английский

Citations

3