| Title:
|
Decoding Stochastic Neural Dynamics Using Explainable Machine Learning: An EEG-Based Candidate Functional Biomarker for Alzheimer’s Disease
|
| Author(s):
|
1
|
| Subject:
|
Artificial Intelligence and Brain Mapping
|
|
|
|
Presenter Author:
|
|
| Name:
|
کوثر فتاحی
|
|
|
|
Author:
|
|
| Name 1:
|
کوثر فتاحی - Corresponding - Presenter
|
| Affiliation 1:
|
National Institute of Genetic Engineering and Biotechnology
|
|
|
| Name 2:
|
|
| Affiliation 2:
|
|
|
|
| Name 3:
|
|
| Affiliation 3:
|
|
|
|
| Name 4:
|
|
| Affiliation 4:
|
|
|
|
| Name 5:
|
|
| Affiliation 5:
|
|
|
|
| Name 6:
|
|
| Affiliation 6:
|
|
|
|
| Name 7:
|
|
| Affiliation 7:
|
|
|
|
| Name 8:
|
|
| Affiliation 8:
|
|
|
|
| Name 9:
|
|
| Affiliation 9:
|
|
|
|
| Name 10:
|
|
| Affiliation 10:
|
|
|
|
Abstract
|
Original Research
|
|
introduction :
|
Early identification of Alzheimer’s disease (AD) is challenging, as standard cognitive assessments often fail to detect subtle preclinical changes. Electroencephalography (EEG) provides a non-invasive measure of neural activity, and recent studies suggest that variability in EEG signals may reflect early disruptions in cortical networks.
|
|
Methods:
|
In this study, multichannel EEG recordings from 16 scalp electrodes following the international 10–20 system were analyzed. Statistical descriptors capturing signal dispersion—including mean, standard deviation, minimum, and maximum—were extracted to characterize neural variability. Supervised machine learning models were trained to predict cognitive status based on these features. Explainable AI techniques, such as SHAP, were applied to assess the relative contribution of individual EEG channels to model predictions.
|
|
Results:
|
The trained models achieved promising classification performance in distinguishing individuals with AD from healthy controls. Interpretability analyses revealed that frontal and parietal electrodes contributed disproportionately to predictive accuracy, aligning with known involvement of executive control and default mode networks in Alzheimer’s disease. The findings indicate that EEG-derived variability measures can capture functionally relevant neural alterations.
|
|
Conclusion:
|
These results support the feasibility of using interpretable machine learning frameworks on EEG data to study neural dysfunction in AD. EEG-based variability measures represent a candidate functional biomarker for early detection, potentially complementing conventional cognitive assessments. Future work should validate these findings in larger, longitudinal cohorts and explore integration with other neuroimaging modalities.
|
| Attachment:
|
|
| Keywords:
|
Alzheimer’s disease, EEG, machine learning, neural variability, explainable AI, functional biomarker
|
| Key message:
|
|
| Do you want your abstract to be indexed, if this is selected as one of the top 100
articles?
|
