What EEG Foundation Models Encode: Dataset Identity and a Negative-Control Suite for Clinical Benchmarks

• URL: http://arxiv.org/abs/2607.24519v2
• Date: Sat, 01 Aug 2026 00:35:45 GMT
• Title: What EEG Foundation Models Encode: Dataset Identity and a Negative-Control Suite for Clinical Benchmarks
• Abstract summary: We benchmark LaBraM, EEGMamba, CBraMod, REVE, LEAD, BENDR, and BIOT on five clinical tasks across four datasets.<n>We challenge apparent gains using stronger classical comparators, label permutation, scrambled-label fine-tuning, and random-initialisation controls.<n>Conclusions change materially after montage matching, patient-overlap checks, stronger comparators, and representation controls.
• Score: 0.0
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Pretrained EEG foundation models are proposed for clinical decoding, but whether reported gains transfer across populations or survive negative controls is unclear. We benchmark LaBraM, EEGMamba, CBraMod, REVE, LEAD, BENDR, and BIOT on five clinical tasks across four datasets. Primary analyses use frozen linear probes with subject-disjoint LOSO or grouped five-fold validation. Because CAUEEG releases no patient identifiers, it is evaluated at recording level with a patient-disjoint sensitivity. We challenge apparent gains using stronger classical comparators, label permutation, scrambled-label fine-tuning, and random-initialisation controls. In a matched 19-channel CAUEEG evaluation (Normal/MCI/Dementia; N = 1,187 recordings), classical features achieve 0.734 macro-AUROC versus 0.699 for BIOT, 0.669 for CBraMod, and 0.568 for REVE. A patient-disjoint sensitivity retains the classical-over-REVE ordering (0.717 versus 0.565). Dataset identity is decoded from frozen REVE embeddings at or near ceiling across Western-Korean and Western-Western pairs, including after PCA-50 and removal of line-frequency and amplitude-scale information. This establishes dataset membership, not a causal site or population effect. A matched random-initialised encoder exceeds pretrained REVE on CAUEEG (0.659 versus 0.570). On CHB-MIT cross-subject ictal detection (n = 23), REVE reaches 0.793, versus 0.739 for the best enhanced nonlinear comparator, 0.701 for random initialisation, and 0.505 for raw-signal random features. Because preprocessing removes absolute amplitude, this does not establish superiority over every plausible handcrafted baseline. Conclusions change materially after montage matching, patient-overlap checks, stronger comparators, and representation controls. We distill these checks into a reporting protocol for clinical EEG foundation-model studies.

Related papers

• A Domain-Structured Ensemble Framework for Perioperative Outcome Prediction Using Electronic Health Record Data [0.0]
  Perioperative risk prediction models are often limited by narrow surgical populations, incomplete data, and limited interpretability.<n>We present a domain-structured ensemble framework for outcome prediction using routinely collected electronic health record (EHR) data.<n>We demonstrate the framework using postoperative delirium in a case-control sample of 5,386 surgical encounters from a statewide health information exchange.
  arXiv  Detail & Related papers  (2026-08-09T21:15:12Z)
• Post-Operative Glioma Segmentation via Loss Stabilization, Normalization and Subspace Attention [46.03321798937855]
  We show that the standard Generalized Dice Loss (GDL) is unstable under domain shift.<n>We propose a Subspace-Aware Class Attention (SACA) module that re-calibrates the bottleneck features and raises Enhancing Tumor (ET) sensitivity by 8%.
  arXiv  Detail & Related papers  (2026-07-23T09:59:30Z)
• Modeling Normal Is All You Need: Joint Latent Clustering for Anomaly Detection in Multimodal Cyber-Physical Systems [47.027290803102666]
  Normal behaviour is the union of many imbalanced, curved, thin-fringed operating regimes.<n>We model the normal law with a jointly learned latent representation plus explicit Gaussian-mixture mode clustering.<n>We evaluate under a deliberately fair protocol: raw point-wise metrics with no point adjustment, a trivial-detector difficulty split, prevalence-matched F1, and train-normal-only calibration.
  arXiv  Detail & Related papers  (2026-07-07T10:10:00Z)
• EEG-SpikeAgent: Agentic Closed-Loop Program Synthesis for Automated EEG Spike Detection [0.0]
  We introduce EEG-SpikeAgent, a closed-loop program-synthesis framework that uses a large language model (LLM) agentic system to generate signal-processing features for spike detection in scalp EEG.<n>Across iterations, EEG-SpikeAgent proposes and refines candidate signal features and decision rules informed by model performance.
  arXiv  Detail & Related papers  (2026-07-06T00:19:41Z)
• TriALS: Triphasic-Aided Liver Lesion Segmentation Benchmark in Non-Contrast CT [49.93685997940356]
  We describe the TriALS challenge for automated liver lesion segmentation under contrast-limited conditions.<n>The top-performing method achieved a mean venous-phase Dice of 0.754, consistent with human-level performance, yet dropped to 0.57 on NCCT.<n>Algorithm performance was most strongly predicted by training data scale and pre-training strategy.
  arXiv  Detail & Related papers  (2026-05-15T19:23:35Z)
• Pretraining Objective Matters in Extreme Low-Data FGVC: A Backbone-Controlled Study [39.146761527401424]
  We compare four frozen ViT-B/16 encoders trained with supervised classification, contrastive learning (SigLIP2), masked reconstruction (MAE), and self-distillation (DINOv3)<n>Supervised and contrastive encoders provide the strongest linear separability (logistic AUC: 0.768 and 0.735; SVM AUC: 0.739 and 0.697)<n>We find that DINOv3 underperforms across probe families in this domain.
  arXiv  Detail & Related papers  (2026-05-15T04:14:16Z)
• Exploring Prompt Alignment with Clinical Factors in Zero-Shot Segmentation VLMs for NSCLC Tumor Segmentation [0.05597620745943382]
  Zero-shot vision-language models (VLMs) offer a promptable alternative to task-specific training for gross tumor volume delineation.<n>We study alignment directions in VoxTell on a held-out internal NSCLC tumor dataset.<n> Alignment analyses revealed that anatomical location is the dominant driver of VoxTell's spatial attention.
  arXiv  Detail & Related papers  (2026-05-02T05:49:32Z)
• Open-Set Vein Biometric Recognition with Deep Metric Learning [0.0]
  We rigorously evaluate the computational boundaries of Deep Metric Learning (DML) under strict open-set constraints.<n>Our approach learns discnative L2-normalised embeddings and employs prototype-based matching with a similarity threshold.<n>On the large-scale MMCBNU 6000 benchmark, our best model (ResNet50-CBAM) achieves an OSCR of 0.9945, AUROC of 0.9974, and EER of 1.57%, maintaining high identification accuracy (99.6% Rank-1) while robustly rejecting unknown subjects.
  arXiv  Detail & Related papers  (2026-04-16T11:03:11Z)
• The Value Sensitivity Gap: How Clinical Large Language Models Respond to Patient Preference Statements in Shared Decision-Making [0.6768558752130311]
  Large language models (LLMs) are entering clinical as decision support tools, yet how they respond to explicit patient value statements remains unmeasured.<n>We conducted a factorial experiment using clinical vignettes derived from 98,759 de-identified Medicaid encounter notes.<n>All models acknowledged patient values in 100% of non-control trials, yet actual recommendation shifting remained modest.
  arXiv  Detail & Related papers  (2026-02-12T21:42:21Z)
• Explainable Admission-Level Predictive Modeling for Prolonged Hospital Stay in Elderly Populations: Challenges in Low- and Middle-Income Countries [65.4286079244589]
  Prolonged length of stay (pLoS) is a significant factor associated with the risk of adverse in-hospital events.<n>We develop and explain a predictive model for pLos using admission-level patient and hospital administrative data.
  arXiv  Detail & Related papers  (2026-01-07T23:35:24Z)
• Knowledge Graph Augmented Large Language Models for Disease Prediction [24.992170033802537]
  Knowledge graph (KG)-guided chain-of-thought (CoT) framework generates clinically grounded reasoning for visit-level disease prediction in MIMIC-III.<n> ICD-9 codes are mapped to PrimeKG, from which disease-relevant nodes and multi-hop reasoning paths are extracted and used as scaffolds for CoT generation.<n> KG-guided models outperform strong classical baselines, achieving AUROC values of 0.66 to 0.70 and macro-AUPR values of 0.40 to 0.47.<n>A blinded clinician evaluation shows consistent preference for KG-guided CoT explanations in clarity, relevance, and clinical correctness.
  arXiv  Detail & Related papers  (2025-12-01T02:49:17Z)
• Transparent Early ICU Mortality Prediction with Clinical Transformer and Per-Case Modality Attribution [42.85462513661566]
  We present a lightweight, transparent multimodal ensemble that fuses physiological time-series measurements with unstructured clinical notes from the first 48 hours of an ICU stay.<n>A logistic regression model combines predictions from two modality-specific models: a bidirectional LSTM for vitals and a finetuned ClinicalModernBERT transformer for notes.<n>On the MIMIC-III benchmark, our late-fusion ensemble improves discrimination over the best single model while maintaining well-calibrated predictions.
  arXiv  Detail & Related papers  (2025-11-19T20:11:49Z)
• I Detect What I Don't Know: Incremental Anomaly Learning with Stochastic Weight Averaging-Gaussian for Oracle-Free Medical Imaging [2.384534878752428]
  We introduce an unsupervised, oracle-free framework that incrementally expands a trusted set of normal samples without any anomaly labels.<n>A frozen pretrained vision backbone is augmented with tiny convolutional adapters, ensuring rapid domain adaptation with negligible computational overhead.<n>On COVID-CXR, ROC-AUC improves from 0.9489 to 0.9982; on Pneumonia CXR, ROC-AUC rises from 0.6834 to 0.8968; and on Brain MRI ND-5, ROC-AUC increases from 0.6041 to 0.7269.
  arXiv  Detail & Related papers  (2025-11-05T23:28:14Z)
• Generalizable Diabetes Risk Stratification via Hybrid Machine Learning Models [0.0]
  Diabetes affects over 537 million people worldwide and is projected to reach 783 million by 2045.<n>We compare two hybrid classifiers and assess their generalizability on an external cohort.
  arXiv  Detail & Related papers  (2025-09-24T21:18:52Z)
• A Novel Attention-Augmented Wavelet YOLO System for Real-time Brain Vessel Segmentation on Transcranial Color-coded Doppler [49.03919553747297]
  We propose an AI-powered, real-time CoW auto-segmentation system capable of efficiently capturing cerebral arteries.<n>No prior studies have explored AI-driven cerebrovascular segmentation using Transcranial Color-coded Doppler (TCCD)<n>The proposed AAW-YOLO demonstrated strong performance in segmenting both ipsilateral and contralateral CoW vessels.
  arXiv  Detail & Related papers  (2025-08-19T14:41:22Z)
• Collaborative residual learners for automatic icd10 prediction using prescribed medications [45.82374977939355]
  We propose a novel collaborative residual learning based model to automatically predict ICD10 codes employing only prescriptions data. We obtain multi-label classification accuracy of 0.71 and 0.57 of average precision, 0.57 and 0.38 of F1-score and 0.73 and 0.44 of accuracy in predicting principal diagnosis for inpatient and outpatient datasets respectively.
  arXiv  Detail & Related papers  (2020-12-16T07:07:27Z)
• Ensemble model for pre-discharge icd10 coding prediction [45.82374977939355]
  We propose an ensemble model incorporating multiple clinical data sources for accurate code predictions. We obtain multi-label classification accuracies of 0.73 and 0.58 for average precision, 0.56 and 0.35 for F1-scores and 0.71 and 0.4 accuracy in predicting principal diagnosis for inpatient and outpatient datasets respectively.
  arXiv  Detail & Related papers  (2020-12-16T07:02:56Z)
• Multilabel 12-Lead Electrocardiogram Classification Using Gradient Boosting Tree Ensemble [64.29529357862955]
  We build an algorithm using gradient boosted tree ensembles fitted on morphology and signal processing features to classify ECG diagnosis. For each lead, we derive features from heart rate variability, PQRST template shape, and the full signal waveform. We join the features of all 12 leads to fit an ensemble of gradient boosting decision trees to predict probabilities of ECG instances belonging to each class.
  arXiv  Detail & Related papers  (2020-10-21T18:11:36Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.