Towards improving Alzheimer's intervention: a machine learning approach for biomarker detection through combining MEG and MRI pipelines
Zur Verbesserung der Alzheimer-Intervention: ein Machine-Learning-Ansatz zur Biomarkerdetektion durch Kombination von MEG- und MRI-Pipelines
Journal of Neural Engineering · 29.10.2026
DOI: 10.1088/1741-2552/aebb66 ↗Auf einen Blick
Bei 324 Personen wurde geprüft, wie gut MEG- und MRI-Merkmale gesunde Kontrollen von Personen mit mild kognitiver Beeinträchtigung unterscheiden. Die beste Leistung erzielte ein kombiniertes Modell mit MRI- und source-basierten MEG-Merkmalen.
Fragestellung
Verbessern MRI-Anatomiemerkmale die Unterscheidung von gesunden Kontrollen und MCI in der BioFIND-Kohorte gegenüber MEG allein?
Methode
Analyse von 324 Personen aus der BioFIND-Kohorte; MEG-Spektralmerkmale auf Sensor- und Source-Ebene, zusätzliche MRI-Merkmale; Vergleich mehrerer Inversionsmethoden und Klassifikatoren mit nested K-fold-Validierung und Monte-Carlo-Repplikationen.
Ergebnisse
Beste Leistung mit MRI plus source-basierten MEG-Merkmalen und GLMNET: 76,31 ± 1,47 % Accuracy und 82,44 ± 1,44 % AUC. MEG-only-Modelle trennten besser als MRI-only-Analysen; SVM und GLMNET waren am konsistentesten effektiv.
Schlussfolgerung der Autor:innen
Die Befunde stützen MEG/EEG-Spektral-Biomarker, besonders source-basierte Darstellungen, als sensible Marker früher AD-bezogener Funktionsänderungen. Die Kombination von MEG mit MRI kann die Klassifikation weiter verbessern.
Kernaussagen
- 324 Personen: 166 HC und 158 MCI
- Beste Klassifikation: MRI + source-basierte MEG-Merkmale mit GLMNET
- MEG-only war stärker als MRI-only; SVM und GLMNET am robustesten
Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.
Kurz gefasst
Fachartikel in der Zeitschrift Journal of Neural Engineering, online erschienen am 29.10.2026. Verfasst von Alwani Liyana Ahmad und 5 weiteren Autor:innen. Beteiligte Einrichtung: Universiti Teknologi PETRONAS, Department of Applied Sciences, Seri Iskandar, 32610, Malaysia. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Ministry of Higher Education (MOHE) Malaysia, Universiti Sains Malaysia, Alberta Innovates LevMax program.
Abstract (Originalsprache)
Objective: To identify robust biomarkers for early cognitive decline, classification performance was evaluated for healthy controls (HC) versus mild cognitive impairment (MCI) using MEG-derived spectral features at the sensor and source levels, and by testing whether adding MRI anatomical features improves discrimination in the BioFIND cohort. Approach: Data from 324 participants (166 HC and 158 MCI) were analyzed using automated artifact rejection and source reconstruction. Spectral biomarkers, including sensor-space and sourcespace representations, were extracted from the magnetometer and gradiometer signals. Five inverse-solution methods were assessed (including linearly constrained minimum variance [LCMV] and exact low-resolution electromagnetic tomography analysis [eLORETA]) to increase the robustness of source-based biomarker selection. Multiple classifiers, including a support vector machine (SVM) and L1-regularized logistic regression (GLMNET), were compared using nested K-fold cross-validation embedded in Monte Carlo replications to limit overfitting. Main results: The best performance was obtained by combining MRI features with source-based MEG features using GLMNET, achieving 76.31 ± 1.47% accuracy and 82.44 ± 1.44% AUC. MEG-only models showed strong discrimination and outperformed MRI-only analyses, particularly those based on the LCMV and eLORETA features. Across settings, the SVM and GLMNET were the most consistently effective classifiers. Significance: These findings support MEG/EEG spectral biomarkers, particularly source-based representations, as sensitive markers of early AD-related functional changes. Integrating MEG with MRI can further improve HC-versus-MCI classification. The study also demonstrated that rigorous validation (nested cross-validation with Monte Carlo replication) is essential for reliable biomarker assessment and for comparing preprocessing, inverse-solution, and classifier choices in translational neuroimaging pipelines.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Universiti Teknologi PETRONAS, Department of Applied Sciences, Seri Iskandar, 32610, Malaysia.
- Ulster University School of Computing Engineering and Intelligent Systems, Ulster University, Derry, Londonderry, BT48 7JL, United Kingdom of Great Britain and Northern Ireland.
- University of Calgary, Department of Radiology and Hotchkiss Brain Institute, Calgary, Alberta, T2N 1N4, Canada.
- Department of Computer Science, University of Bath, Bath Institute for the Augmented Human, Bath, Bath, BA2 7AY, United Kingdom of Great Britain and Northern Ireland.
- Hospital Pakar Universiti Sains Malaysia, Department of Neurosciences, Kota Bharu, Kelantan, 11800, Malaysia.
ORCID-Profile: José Miguel Sánchez-Bornot, Roberto C Sotero, Damien H Coyle, Zamzuri Idris
Förderung
- Ministry of Higher Education (MOHE) Malaysia
- Universiti Sains Malaysia
- Alberta Innovates LevMax program
- Medical Research Council
- MRC Dementias Platform UK
Zitieren
APA
Ahmad, A. L., Sánchez-Bornot, J. M., Sotero, R. C., Coyle, D. H., Idris, Z., & Faye, I. (2026). Towards improving Alzheimer's intervention: a machine learning approach for biomarker detection through combining MEG and MRI pipelines. Journal of Neural Engineering. https://doi.org/10.1088/1741-2552/aebb66
Vancouver
Ahmad AL, Sánchez-Bornot JM, Sotero RC, Coyle DH, Idris Z, Faye I. Towards improving Alzheimer's intervention: a machine learning approach for biomarker detection through combining MEG and MRI pipelines. Journal of Neural Engineering. 2026. doi:10.1088/1741-2552/aebb66
BibTeX
@article{Ahmad2026,
author = {Ahmad, Alwani Liyana and Sánchez-Bornot, José Miguel and Sotero, Roberto C and Coyle, Damien H and Idris, Zamzuri and Faye, Ibrahima},
title = {Towards improving Alzheimer's intervention: a machine learning approach for biomarker detection through combining MEG and MRI pipelines},
journal = {Journal of Neural Engineering},
year = {2026},
doi = {10.1088/1741-2552/aebb66}
}
EUVIMED