EEG models show varied stability across different electrode setups
Representational and Functional Robustness to Electrode Montages in EEG Foundation Models
Machine Learning
Summary
EEG brain signal models are meant to work well even when the electrodes recording brain activity are placed differently. The authors studied how four such models handled changes in electrode layouts and found that some models stay stable in their internal signals but still perform worse in tasks, while others behave differently. They showed that how the models combine information from electrodes plays a big role in their stability. This means that simply accepting different electrode layouts is not enough; both the model and its way of summarizing data must be robust.
What this means in practice
- •For medical device developers: Design EEG systems that remain accurate despite varying electrode placements by improving model and data aggregation robustness.
- •For brain-computer interface engineers: Build more reliable motor imagery classifiers by choosing EEG models and aggregation methods proven to maintain functionality across electrode setups.
Authors
Jakob Steglich, Justus Meyer zu Bexten, Shakiba Moradi, Laure Ciernik, Simon M. Hofmann, Mina Jamshidi Idaji
Abstract
EEG foundation models (EEG-FMs) are intended to generalize across different datasets by learning representations that, ideally, are invariant to dataset-specific EEG configurations such as electrode montages. However, EEG-FMs that accept different montages as input do not guarantee that representations and predictions remain stable across different electrode configurations, especially outside the training setting. In this work, we investigate the effects of different electrode montages through a joint functional and representational analysis of four EEG foundation models selected to span distinct montage-handling designs. We evaluate embeddings on cross-subject resting-state eyes-open/closed and within-subject motor-imagery classification under spatially informed channel reduction. Functional robustness is tested through the generalizability of linear probes across channel counts, while representational robustness is assessed through within-subject similarity and preservation of between-subject geometry. The four models show distinct robustness profiles, and the two axes dissociate: large changes in embedding similarity need not come with comparable probe degradation, and stable embeddings can still lose downstream performance. Comparing two readouts of the same encoder further shows that aggregation, not the encoder alone, determines functional robustness: pooling into anatomically aligned regions degrades less than a learned global readout, despite being montage-invariant by construction. Montage robustness is therefore a joint property of the encoder and its aggregation, and characterizing it requires both a representational and a functional axis. Input compatibility alone is evidence for neither.