Enhanced Alzheimer’s detection with EEG source imaging and multi-branch joint attention
Bibliographic record
Abstract
Abstract Objective. Alzheimer’s disease (AD) is a neurodegenerative disorder detectable via electroencephalogram (EEG). Traditional EEG-based AD detection methods do not fully leverage spatial information about brain activity and the correlation between different frequency bands, leading to suboptimal cross-patient performance. Approach. We propose a new multi-branch joint attention network (MJANet) based on electrophysiological source imaging (ESI) to create comprehensive spatial power maps, improving the spatial resolution of EEG. The MJANet incorporates a multi-branch joint attention (MBJA) mechanism to capture interactions across different frequency bands. The MJANet employs a new MBJA mechanism to capture interactions across frequency bands and a moving shifted window to capture global image features. It analyzes correlations between activities in various bands and brain regions to boost cross-patient detection capabilities. Main results. The proposed approach is validated on a public dataset with a leave-one-subject-out cross-validation strategy, achieving an 85.23% accuracy rate in differentiating AD from normal controls (NC), representing an 8.03% improvement over the state-of-the-art. Moreover, it achieves 75.57% accuracy in distinguishing frontotemporal dementia (FTD) from NC, and 63.97% for the classification of AD, NC, and FTD. We utilize GradCAM to visualize the joint attention mechanism, providing insights into its decision-making process. Significance. This work explores a novel biomarker that has the potential to enhance clinical diagnostic methods and improve diagnostic accuracy.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".