Skip to content

A dual-branch time–frequency fusion network for EEG-based classification of Alzheimer's disease and frontotemporal dementia

Sep 2026 · Journal of Alzheimer's Disease · Vol 113, pp. 1857 - 1871 · 0 citations · 43 references
Medicine

TL;DR

The proposed DBTF-Net leverages temporal and time-frequency information in EEG signals and provides classification of AD and FTD, and visualization analysis indicates that the model attends to disease-relevant discriminative patterns in time-frequency representations, enhancing the interpretability of its classification decisions.

Abstract

Background Alzheimer's disease (AD) and frontotemporal dementia (FTD) exhibit substantial overlap in clinical manifestations and patterns of brain functional degeneration, which poses significant challenges for automated classification based on electroencephalography (EEG). Objective This study aims to develop an EEG-based framework capable of simultaneously capturing temporal dynamics and frequency-related characteristics of EEG signals for discrimination among AD, FTD, and cognitively normal (CN) subjects. Methods A Dual-Branch Time-Frequency Fusion Network (DBTF-Net) based on routine clinical resting-state EEG recordings acquired under eyes-closed conditions is proposed. The model employs parallel temporal and frequency branches to process raw EEG time-series signals and their corresponding time-frequency representations. A global temporal dependency construction mechanism is introduced in the temporal branch to capture both local temporal patterns and long-range temporal dependencies. Feature-level fusion is then performed across the two branches to achieve a collaborative representation of multidimensional brain functional information. The proposed method was systematically evaluated on one three-class classification task (AD versus FTD versus CN) and multiple binary classification tasks. Results Experimental results from five-fold cross-validation at the epoch level show the classification accuracies of DBTF-Net as 86.36%±4.28%, 83.01%±6.15%, 92.13%±10.35%, and 88.74%±7.69%% for AD versus FTD versus CN, AD versus CN, FTD versus CN, and AD versus FTD, respectively. Conclusions The proposed DBTF-Net leverages temporal and time-frequency information in EEG signals and provides classification of AD and FTD. Visualization analysis further indicates that the model attends to disease-relevant discriminative patterns in time-frequency representations, enhancing the interpretability of its classification decisions.

View source

Similar papers

Aug 2026

EEG Time–Frequency Image Encoding with Convolutional Networks for Alzheimer’s and FTD Classification

This study presents a deep learning framework that leverages EEG-derived time–frequency images and convolutional neural networks (CNNs) for robust, non-invasive classification of dementia subtypes, and consistently outperforms all baseline methods across accuracy, precision, recall, and F1-score.

Salleh Sonko, Gahangir Hossain, Wael Korani · 1 citation
Conference Open access Aug 2026

Early Diagnosis and Classification of Alzheimer’s Disease Based on EEG Time–Frequency Images

The proposed intelligent EEG diagnostic framework shows potential for deployment in primary healthcare institutions and may provide theoretical support for addressing the growing challenges of AD diagnosis and treatment in the context of global population aging.

Yuxuan Qi · 0 citations
Open access Aug 2026

Interpretable AI for neural signal decoding in dementia: an EEG ensemble approach to differential diagnosis

A domain-informed heterogeneous ensemble framework incorporating dynamic neural reactivity and hemispheric asymmetry metrics from 19-channel EEG recordings acquired from 88 participants indicates that the integration of dynamic state-transition measures with structural asymmetry proxies enhances electrophysiological di...

Fawad Muhammad, I. Usmani, M. Aamir et al. · 0 citations
Open access Aug 2026

Spectral and Directed Connectivity EEG Markers for Classifying Alzheimer’s Disease, Frontotemporal Dementia, and Healthy Controls with Exploratory Photobiomodulation Case-Study Projection

Alzheimer’s disease (AD) and frontotemporal dementia (FTD) are neurodegenerative disorders with partially overlapping clinical manifestations, making early and differential diagnosis challenging. This study investigated whether electroencephalography (EEG)-derived spectral features and Granger-causality (GC)-based dire...

Z. Šverko, S. Vlahinić, M. Vrankić et al. · 0 citations
Open access Sep 2026

An Interpretable EEG Machine Learning Prototype for Alzheimer’s Disease Classification

An interpretable, subject-level AD versus healthy-control classifier from the dataset OpenNeuro ds004504 is developed and deployed as a public research prototype, not a clinical diagnostic device, and requires external validation before any clinical interpretation.

Atrim Das, A. Chakravarty · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.