EEG-Based Classification of Parkinson’s Disease With Freezing of Gait Using Midfrontal Beta Oscillations
Journal of Integrative Neuroscience · 20.06.2025 · 24 (6)
DOI: 10.31083/jin39023 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Integrative Neuroscience, online erschienen am 20.06.2025. Verfasst von Shotabdi Roy und 7 weiteren Autor:innen. Beteiligte Einrichtung: Biomedical and Translational Sciences, University of South Dakota, Vermillion, SD 57069, USA. Hauptthemen laut MEDLINE: Parkinson Disease, Beta Rhythm, Gait Disorders, Neurologic, Electroencephalography. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 44 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: SSOM Biomedical AI and Computation Idea Competition (BACIC).
Abstract (Originalsprache)
Background
Freezing of gait (FOG) is a debilitating motor symptom of Parkinson’s disease (PD) that significantly affects patient mobility and quality of life. Identifying reliable biomarkers to distinguish between PD patients with freezing of gait (PDFOG+) and those without FOG (PDFOG–) is essential for early intervention and treatment planning. This study investigates the potential of electroencephalographic (EEG) signals, focusing on well-studied midfrontal beta oscillatory feature, to classify PDFOG+ and PDFOG– using machine learning (ML) and deep learning (DL) approaches.
Methods
Resting-state EEG data were collected from the midfrontal ‘Cz’ and nearby channels (Cz-cluster) from 41 PDFOG+ and 41 PDFOG– subjects. A range of ML and DL models, including logistic regression (LR), random forest (RF), extreme gradient boosting (XGBoost), categorical boosting (CatBoost), and long short-term memory (LSTM) models were evaluated using leave-one-subject-out (LOSO), 10-fold, and stratified cross-validation (CV).
Results
Outcomes demonstrate that while LR achieved an area under the receiver-operating characteristic (AUC-ROC) score of 0.63, LSTM outperformed all models, achieving an AUC-ROC of 0.68 and accuracy of 0.63, particularly with the Cz-cluster configuration.
Conclusions
These findings support the potential of midfrontal beta oscillations, particularly in combination with LSTM temporal modeling, a promising EEG-based biomarker for distinguishing PDFOG+ from PDFOG–. This work contributes to the development of more effective diagnostic tools and treatment strategies for PD-related gait impairments.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Biomedical and Translational Sciences, University of South Dakota, Vermillion, SD 57069, USA.
- Neuroergonomics and Cognitive Engineering Lab, Oklahoma State University, Stillwater, OK 74078, USA.
- Department of Psychology, University of South Dakota, Vermillion, SD 57069, USA.
- Department of Neurology, Ludwig Maximilian University, 81377 Munich, Germany.
- Applied AI Research Lab, Department of Computer Science, The University of South Dakota, Vermillion, SD 57069, USA.
- Department of Neuroscience, Sanford School of Medicine, University of South Dakota, Sioux Falls, SD 57104, USA.
Förderung
- SSOM Biomedical AI and Computation Idea Competition (BACIC)
Zitieren
APA
Roy, S., Nuamah, J., Bosch, T. J., Barsainya, R., Scherer, M., Koeglsperger, T., Santosh, K. C., & Singh, A. (2025). EEG-Based Classification of Parkinson’s Disease With Freezing of Gait Using Midfrontal Beta Oscillations. Journal of Integrative Neuroscience, 24(6). https://doi.org/10.31083/jin39023
Vancouver
Roy S, Nuamah J, Bosch TJ, Barsainya R, Scherer M, Koeglsperger T, et al. EEG-Based Classification of Parkinson’s Disease With Freezing of Gait Using Midfrontal Beta Oscillations. Journal of Integrative Neuroscience. 2025;24(6). doi:10.31083/jin39023
BibTeX
@article{Roy2025,
author = {Roy, Shotabdi and Nuamah, Joseph and Bosch, Taylor J and Barsainya, Richa and Scherer, Maximilian and Koeglsperger, Thomas and Santosh, K C and Singh, Arun},
title = {EEG-Based Classification of Parkinson’s Disease With Freezing of Gait Using Midfrontal Beta Oscillations},
journal = {Journal of Integrative Neuroscience},
year = {2025},
volume = {24},
number = {6},
doi = {10.31083/jin39023}
}
EUVIMED