Papers for

bci system developers

Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.

Adaptive time window selection improves motor imagery brain interface accuracy

Time Window Optimization for Motor Imagery Classification via Temporal Structure Embedded Reinforcement Learning

Abstract: Online motor imagery-based brain-computer interfaces (MI-BCIs) typically decode continuous EEG using fixed-length sliding windows that may be misaligned with neural events. Existing window optimization methods often evaluate windows independently, overlooking temporal dependencies and potentially producing fragmented or physiologically implausible selections. We propose Temporal Structure Embedded Reinforcement Learning (TSE-RL), a framework that incorporates prior knowledge of MI event structure into a constrained Markov decision process for adaptive window optimization. First, TSE-RL assigns each window an explicit semantic role through Discard, Onset, Sustain, or Offset actions, extending binary retain/discard decisions to event-level alignment. Second, a finite-state machine masks invalid actions to enforce the temporal order of MI phases and promote physiologically plausible window sequences. Third, a Transformer-based history encoder captures dependencies across windows, improving training stability and computational efficiency compared with recurrent architectures. TSE-RL is plug-and-play and can be integrated with diverse MI classifiers. Experiments on multiple benchmark datasets demonstrate consistent improvements in decoding performance in both cross-session and cross-subject settings. Visualization analyses further suggest that the learned window selection policy aligns with neurophysiological principles.

Mon 12 OctNeural and Evolutionary ComputingMachine Learning
The gist
Motor imagery brain-computer interfaces read brain signals to control devices, but they usually use fixed time windows that don’t always match the brain’s natural activity. The authors propose a new method that picks time windows more intelligently by recognizing different phases of brain activity and keeping the sequence realistic. Their approach uses reinforcement learning with a model that understands the order of brain events and remembers past windows better. This method works well across different people and sessions, making brain signal decoding more accurate.
Open → 2610.16921v1

Covariance transform reduces identity risk in motor imagery eeg data

Riemannian Suppression of Cross-Session Identity Leakage in Motor-Imagery EEG

Abstract: Electroencephalography (EEG) recorded for motor-imagery (MI) brain-computer interfaces (BCIs) carries subject-identifying structure that persists across sessions, allowing separately released sessions from the same person to be linked. Existing EEG identity protections use learned or additive perturbations and operate on representations other than the covariance geometry underlying MI pipelines. We ask whether cross-session identity leakage can be suppressed directly in this covariance representation without learning a privacy model. We introduce a Riemannian transform that equalizes two subject-conditioned geometric statistics: the Riemannian mean covariance and the tangent-space scatter around it, with two optional band-specific refinements. Ablations show that mean alignment and scatter whitening carry nearly all of the suppression, while the optional stages contribute small, dataset-dependent refinements. Across two public datasets and five attackers spanning four covariance-space representations, the worst attacker falls from 90.8% to 15.6% on BCI Competition IV-2a and from 90.8% to 9.9% on OpenBMI, at a cost of 7.0 and 2.1 points of MI decoding accuracy relative to the 8-30 Hz input, and no aggregation rule exceeds chance by more than two points at any batch size. The transform requires no learned privacy model, surrogate network, or gradient-based training, and is fitted on one session and applied unchanged to a held-out later session.

Sun 11 OctCryptography and Security
The gist
Motor-imagery EEG signals can reveal who a person is, even across multiple recording sessions, which raises privacy concerns. The authors developed a mathematical method that adjusts the core features of the EEG data to hide these personal identifiers without needing complex training or guessing models. Their technique greatly lowers the chance that someone can identify a person from their EEG data across sessions, while still keeping the important signals needed for brain-computer interface tasks mostly intact. This method was tested on public datasets and worked against various methods trying to reveal identities.
Open → 2610.15626v1

Cortical alignment improves eeg model accuracy across montages

CortexBridge: Cortical Alignment of EEG Montages for Foundation Models

Abstract: Electroencephalography (EEG) foundation models are often pretrained with a fixed channel vocabulary or a limited set of montages, making transfer difficult when electrode layouts change. We propose CortexBridge, a lightweight adapter that combines EEG features with electrode and atlas coordinates to map arbitrary montages into a shared cortical latent space. Evaluated with three frozen foundation models on five brain-computer interface (BCI) datasets from the Mother of All BCI Benchmarks (MOABB), CortexBridge improves performance in 13 of 15 evaluations. The gains in balanced accuracy average 0.80% for EEGPT, 0.70% for LaBraM, and 3.26% for CBraMod, with a maximum gain of 13.02% on 12-class steady-state visual evoked potential (SSVEP) classification. Visualizations of the learned atlas representations reveal task-dependent spatial patterns, with SSVEP showing a more concentrated representation in the Yeo Visual network than auditory P300. These results establish cortical alignment as a learnable and anatomically grounded routing mechanism from heterogeneous EEG montages to pretrained foundation models.

Thu 1 OctArtificial Intelligence
The gist
EEG devices measure brain activity using different layouts of electrodes, which makes it hard for computer models trained on one layout to work well with others. The authors propose CortexBridge, a small addition that aligns these different layouts to a common brain map, letting existing models understand data from varied electrode setups. Testing on several brain-computer interface tasks showed CortexBridge generally improved accuracy. The authors also found that the method can reveal meaningful brain activity patterns depending on the task.
Open → 2610.01124v1

AutoBCI finds better brain computer interface designs using AI agents

AutoBCI: Forecast-Guided Agentic Neural Architecture Discovery for EEG-Based Brain--Computer Interfaces

Abstract: EEG-based brain-computer interfaces support a broad range of applications, yet designing decoding architectures that perform well across diverse tasks remains challenging. We introduce AutoBCI, an agentic framework in which a Designer Agent and a Forecaster Agent support the discovery and selection of EEG decoding architectures across tasks. The Designer Agent performs Pool-Guided Architecture Discovery (PGAD), generating and refining architectures through training and validation across multiple EEG tasks, such as emotion recognition, motor imagery, and sleep staging. The Forecaster Agent performs Performance Estimation from Early Knowledge (PEEK), using architecture code, the training protocol, and early learning curves to predict full-budget validation performance and select promising candidates for continued training. Across 14 EEG datasets spanning motor imagery, emotion recognition, and sleep staging, we evaluate AutoBCI with six LLMs, including Opus 5.5 and GPT 5.6 Sol, and compare the architectures selected by the search procedure against ten baselines: six conventional EEG models and four foundation models. The architecture discovered by AutoBCI with Claude Opus 5.5 achieves 64.16% average test balanced accuracy (bAcc), compared with 63.87% for REVE, the strongest baseline on this metric. Using ten observed epochs, PEEK reduces mean absolute error in predicting average validation bAcc from 2.20 to 1.36 percentage points, a 38.1% reduction relative to the best-observed-score baseline.

Mon 28 SeptArtificial Intelligence
The gist
Brain computer interfaces (BCIs) read brain signals to control computers, but designing the software for this is hard due to many tasks. The authors created AutoBCI, a system where two AI agents work together: one designs brain signal decoding methods and the other guesses how well these methods will perform early on. This helps find effective designs faster and across different brain-related tasks like recognizing emotions or sleep states. Their approach slightly outperforms existing models on various test sets.
Open → 2609.35456v1

Personalizing EEG models boosts accuracy beyond general brain data gains

Separating personal from population gains when calibrating EEG foundation models for new users

Abstract: Foundation models are increasingly adapted to individual users, but an apparent personalization gain can simply reflect a stronger population model. This distinction matters for brain-computer interfaces, where every new user must be calibrated. We evaluated personal adaptation of three frozen EEG foundation models (CBraMod, REVE and LaBraM) in 235 held-out subjects from three motor-imagery datasets, comparing each subject's adapter with the population model and with adapters fitted to other subjects. Using all first-half session labels, personal adapters improved mean balanced accuracy over the population model by 1.5-5.4 percentage points and outperformed exchanged adapters by 2.3-7.3 points in all nine model-dataset combinations. The size of this benefit depended on population training: with four times the original budget, median gains remained positive (1.0-2.0 points) but were smaller for every model, and no population model reached a confirmed plateau. Acquiring the benefit cheaply was unreliable: few-label calibration was consistently non-negative on only one dataset, and in CBraMod neither unlabeled context nor meta-learned initialization outperformed matched controls. Personalization should therefore be evaluated against both a population reference and exchanged parameters, across population-training budgets.

Mon 28 SeptMachine Learning
The gist
Calibrating brain-computer interface (BCI) models for each new user can make them work better, but sometimes improvements are just because the overall model is better, not the personal tuning. The authors tested three advanced EEG models on hundreds of new users and showed that tuning for each person really does improve performance over general models or using other people's tuning. However, the size of this personal benefit depends on how much population data was used to train the base model, and using few labels or unlabeled data isn’t always reliable for personalization. They suggest comparing personal tuning results against both the base model and tuning from other users to truly measure gains.
Open → 2609.34801v1