EEG Pathology Detection via Clinical Concept Bottleneck Models

Topic: Brain imaging, connectivity and assessment

A bstract Deep learning models for electroencephalography (EEG) pathology detection achieve competitive accuracy but offer limited interpretability. In this paper, we introduce concept bottleneck models (CBMs) for EEG pathology detection which use clinical EEG biomarkers computed analytically from the signal as a bottleneck to be used for prediction. Those concepts include band power, spectral ratios, alpha peak frequency and amplitude, hemispheric asymmetry and phase locking values. With the backbone based on shallow convolutional neural network (CNN), we compare a black box baseline with two variants of CBMs. The hybrid variant is a classifier receives a residual projection of the backbone features. The pure variant is a CBM where the classifier receives only the predicted concepts. Models are evaluated on abnormality detection and dementia classification followed by concept prediction quality and leakage analysis. Both CBMs approach the baseline on TUAB (82.7–82.9% vs 83.6%) and are 2–5 percentage points lower on CAUEEG (57.9–60.5% vs 62.7%). The pure CBM exhibits concept fidelity concern where the predicted concepts deviate from their ground truth values. Replacing them with their true values causes severe cost on accuracy by 26 and 22 points on TUAB and CAUEEG. The concept fidelity concern is mitigated by increasing the weight of the concept loss to match the ground truth concepts. the mitigation comes with small accuracy cost on TUAB but a relatively larger cost on CAUEEG. The code of the experiments and full logs are available at https://www.github.com/mhmdrdwn/cbm .

Key findings

• Both CBMs approach the baseline on TUAB (82.7–82.9% vs 83.6%) and are 2–5 percentage points lower on CAUEEG (57.9–60.5% vs 62.7%).
• The pure CBM exhibits concept fidelity concern where the predicted concepts deviate from their ground truth values.
• Replacing them with their true values causes severe cost on accuracy by 26 and 22 points on TUAB and CAUEEG.
• The concept fidelity concern is mitigated by increasing the weight of the concept loss to match the ground truth concepts. the mitigation comes with small accuracy cost on TUAB but a relatively larger cost on CAUEEG.
• The code of the experiments and full logs are available at https://www.github.com/mhmdrdwn/cbm .

New to this subject? Start with our plain-English guide: Digital health →

Study details

• Effect size82.9%; 83.6%; 60.5%
• Evidence strengthLow

Read the original

• DOI: 10.64898/2026.10.04.756607
• Original source →

Free copy unknown: no free-copy check has an answer for this record yet. That does not mean it is paywalled.

@article{radwan2026-cccc408d,
  author  = {Radwan, M. and Lind, P. G. and Yazidi, A.},
  title   = {{EEG Pathology Detection via Clinical Concept Bottleneck Models}},
  journal = {bioRxiv},
  year    = {2026},
  doi     = {10.64898/2026.10.04.756607},
  url     = {https://doi.org/10.64898/2026.10.04.756607}
}

Related studies in Digital health tools

• Alzheimer's Disease Detection via Large Pre-Trained Speech and Language Models: Taxonomy, Progress, and Open Challenges PsyArXiv · 2026-09-03 · usefulness score 16/100
• Amelioration of Alzheimer's disease pathology by mitophagy inducers identified via machine learning and a cross-species workflow Nature biomedical engineering · 2022-01-06 · usefulness score 43/100
• A Single-Channel EEG-Based Approach to Detect Mild Cognitive Impairment via Speech-Evoked Brain Responses IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society · 2019-04-18 · usefulness score 48/100
• A New Framework for Automatic Detection of Patients With Mild Cognitive Impairment Using Resting-State EEG Signals IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society · 2020-07-31 · usefulness score 46/100
• DICE-Net: A Novel Convolution-Transformer Architecture for Alzheimer Detection in EEG Signals IEEE Access · 2023-01-01 · usefulness score 45/100
• Standardizing Number of EEG Sensors for AI-Driven Dementia Detection IEEE Sensors Letters · 2025-04-14 · usefulness score 38/100

Also tagged “convolutional”

205 studies in the library carry the topic tag “convolutional”; these are this study's neighbours in a fixed, alphabetical-by-hash tour of them, not the closest in content. All 205 tagged studies.

• Brain MRI analysis for Alzheimer's disease diagnosis using an ensemble system of deep convolutional neural networks Brain informatics · 2018-05-31 · usefulness score 41/100 · Digital health tools
• Hippocampal surface analysis using discrete Ricci flow and Riemannian metrics for the diagnosis of Alzheimer's disease Research Square · 2026-09-18 · usefulness score 16/100 · Getting a diagnosis
• Predicting cognitive decline with deep learning of brain metabolism and amyloid imaging Behavioural brain research · 2018-02-14 · usefulness score 66/100 · Digital health tools

Similar studies

• Interpretable Frequency-Aware Deep Learning for Multi-Class EEG Classification of Neurological Disorders 2026 · Brain scans and imaging
• Ensembles of Deep Learning Architectures for the Early Diagnosis of the Alzheimer's Disease 2016 · Digital health tools
• A Dual-Stream Spatio-Temporal Network with Bidirectional Rectification for Differential Diagnosis of Dementia using EEG 2026 · Brain scans and imaging
• Deep learning-based feature representation for AD/MCI classification 2013 · Digital health tools
• DEMNET: A Deep Learning Model for Early Diagnosis of Alzheimer Diseases and Dementia From MR Images 2021 · Long-term cohort studies
• Predictive EEG Signatures of Alzheimer's Disease: A Cross-Validated, Cross-Method Analysis Using ANCOVA Filtering and Sparse Logistic Regression 2026 · Brain scans and imaging

More on Digital health tools

• A review that covers this ground Interpreting artificial intelligence models: a systematic review on the application of LIME and SHAP in Alzheimer's disease detection Brain informatics · 2024-04-05 · usefulness score 72/100
• Newest in this subject Speaker-Wise Dementia Screening fromNaturalistic Speech with Lexical Entropy, Semantic Drift, and Multimodal Fusion Research Square · 2026-10-21 · Preprint — not yet peer-reviewed · usefulness score 16/100
• Newest in this subject Integrating Multisource Target Evidence with Machine-Learning Potency Prediction for Alzheimer’s Disease Drug Discovery bioRxiv · 2026-10-20 · Preprint — not yet peer-reviewed · usefulness score 16/100

See all studies in Digital health tools →