# Rethinking EEG-Based Disease Diagnosis: Decoupling Instance Representation Learning from Subject-Level Supervision

arXiv: https://arxiv.org/abs/2607.27274 · HTML: https://mlsanity.com/paper/2607.27274/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 12 of its 31 references that are in the map.
- First posted on arXiv 29 Jul 2026.

## Builds on (most cited first, 12 of 12)

- [Attention-based Deep Multiple Instance Learning](https://mlsanity.com/paper/1802.04712/) (2018)
- [EEGNet: A Compact Convolutional Network for EEG-based Brain-Computer Interfaces](https://mlsanity.com/paper/1611.08024/) (2016)
- [Dual-stream Multiple Instance Learning Network for Whole Slide Image Classification with Self-supervised Contrastive Learning](https://mlsanity.com/paper/2011.08939/) (2020)
- [A Transformer-based Framework for Multivariate Time Series Representation Learning](https://mlsanity.com/paper/2010.02803/) (2020)
- [Large Brain Model for Learning Generic Representations with Tremendous EEG Data in BCI](https://mlsanity.com/paper/2405.18765/) (2024)
- [CBraMod: A Criss-Cross Brain Foundation Model for EEG Decoding](https://mlsanity.com/paper/2412.07236/) (2024)
- [Revisiting Multiple Instance Neural Networks](https://mlsanity.com/paper/1610.02501/) (2016)
- [REVE: A Foundation Model for EEG -- Adapting to Any Setup with Large-Scale Pretraining on 25,000 Subjects](https://mlsanity.com/paper/2510.21585/) (2025)
- [Contrast Everything: A Hierarchical Contrastive Framework for Medical Time-Series](https://mlsanity.com/paper/2310.14017/) (2023)
- [Additive MIL: Intrinsically Interpretable Multiple Instance Learning for Pathology](https://mlsanity.com/paper/2206.01794/) (2022)
- [MBrain: A Multi-channel Self-Supervised Learning Framework for Brain Signals](https://mlsanity.com/paper/2306.13102/) (2023)
- [DSAINet: An Efficient Dual-Scale Attentive Interaction Network for General EEG Decoding](https://mlsanity.com/paper/2604.18095/) (2026)

## Abstract (arXiv)

EEG-based disease diagnosis requires one prediction per subject, yet common pipelines segment recordings into short instances, inherit the subject label for every instance, and train instance-level classifiers. This assumes that all instances provide equally reliable diagnostic evidence. Multiple instance learning (MIL) avoids inherited labels by treating each subject as a bag. However, EEG datasets contain far fewer subjects than instances, which can limit the quality of the representations learned by end-to-end MIL. We propose BridgeMIL, a two-stage framework that decouples instance representation learning from subject-level supervision. Stage 1 pretrains the encoder without inherited instance labels by aligning temporally nearby windows and independently sampled within-subject sub-bags. Variance and covariance regularization prevent collapse and reduce redundancy without negative pairs. Stage 2 transfers the encoder to an attention-based MIL aggregator, applies supervision only to subject predictions, and limits representation drift through feature retention. Across three EEG disease datasets and five representative backbones, BridgeMIL attains the highest mean accuracy in 14 of 15 dataset-backbone settings and an overall mean accuracy of 76.57%, 4.28 percentage points higher than the strongest baseline. Further analyses reveal substantial variation in inherited-label reliability across instances, greater performance sensitivity to subject scarcity than to instance scarcity, and a more structured representation space with distinct subject-wise clusters and improved separation between diagnostic classes. Together, these findings underscore the importance of aligning supervision with the subject-level prediction objective while learning from abundant EEG instances without assigning disease labels to individual instances.