Deep Model Families for EEG-Based Multiclass Dementia Classification
- 3 July 2026
- journal article
- Published by World Scientific Pub Co Pte Ltd in International Journal of Neural Systems
- Vol. 37 (02) , 2750005
Abstract
Deep learning (DL) has shown considerable promise for EEG-based dementia assessment; however, rigorous cross-family comparisons under leakage-free and clinically meaningful evaluation protocols remain limited. To address this gap, we benchmarked twelve DL architectures from four major families—Recurrent Neural Networks (RNNs), Temporal Convolutional Networks (TCNs), Transformers, and Self-Supervised Learning (SSL)—for the multi-class differentiation of Alzheimer’s disease (AD), Frontotemporal Dementia (FTD), and Healthy Controls (HC). The analysis was conducted on a clinically acquired EEG dataset comprising 36 AD patients, 23 FTD patients, and 29 HC participants. To prevent subject-level data leakage, we employed strict subject-independent 5-fold cross-validation, and to improve clinical relevance, final predictions were aggregated at the subject level using majority voting across EEG segments. Model performance was evaluated using classification metrics, learning stability, statistical ranking, computational efficiency, and interpretability analysis. Among all benchmarked architectures, the self-supervised SimCLR-Time framework achieved the strongest subject-level performance, reaching a mean accuracy of [ ] and a weighted F1-score of [ ]. The Time-Series Transformer (TST) was the best-performing fully supervised model, with an accuracy of [ ], followed closely by the hybrid LSTM-FCN model at [ ]. In contrast, the EEG-Seq2Seq model showed the weakest performance [ ]. Computational profiling further showed that TCN-Attention offered a favorable trade-off between accuracy and edge-oriented efficiency, while SimCLR-Time combined strong diagnostic performance with the fastest inference latency. In addition, Integrated Gradients-based saliency analysis suggested that the top-performing model relied on structured spatiotemporal EEG patterns rather than arbitrary temporal artifacts. Overall, these findings provide a controlled within-dataset benchmark for multi-class dementia classification from raw EEG and suggest that self-supervised and attention-based architectures are promising directions for clinically oriented EEG decision-support systems. However, external validation on larger and more heterogeneous cohorts remains necessary before broader generalization can be claimed.Keywords
Funding Information
- Izmir University of Economics, Scientific Research Projects Coordination Unit (2022-07)
This publication has 40 references indexed in Scilit:
- Supervised machine learning methods for short-term prediction of a sudden cardiac death from electrocardiogramBiocybernetics and Biomedical Engineering, 2026
- A contrastive learning model of attentional enhancement for EEG cross-subject emotion recognitionBiomedical Signal Processing and Control, 2025
- Detection of hyperglycemia and hypoglycemia using deep learning from facial images obtained with an AI image generatorBiomedical Signal Processing and Control, 2025
- Classification for Alzheimer's disease and frontotemporal dementia via resting-state electroencephalography-based coherence and convolutional neural networkCognitive Neurodynamics, 2025
- The Time-Sequence Prediction via Temporal and Contextual Contrastive Representation Learning2022
- A deep learning based model using RNN-LSTM for the Detection of Schizophrenia from EEG dataComputers in Biology and Medicine, 2022
- An FCN-LSTM model for neurological status detection from non-invasive multivariate sensor dataNeural Computing and Applications, 2022
- Machine learning techniques for diagnosis of alzheimer disease, mild cognitive disorder, and other types of dementiaBiomedical Signal Processing and Control, 2022
- Electroencephalography-based motor imagery classification using temporal convolutional network fusionBiomedical Signal Processing and Control, 2021
- A New dispersion entropy and fuzzy logic system methodology for automated classification of dementia stages using electroencephalogramsClinical Neurology and Neurosurgery, 2021