Decoding Algorithms

Motor imagery BCIs have long faced two problems: wide variation between users and only a small number of distinguishable commands. In this study, users first select a target by looking at it, then confirm the choice with imagined movement, merging the two steps into one. In tests with 15 healthy participants using 16-channel EEG, the hybrid paradigm outperformed motor imagery alone in every channel configuration, reaching up to 100% accuracy. The researchers also found that fixating on the target made EEG responses more stable, which supports using fewer electrodes and lowering the hardware barrier.

Phase mismatch between live EEG and fixed templates has long been the weak point of asynchronous steady-state visual evoked potential (SSVEP) brain-computer interfaces, which are seen as a promising route to real-world control. A new method, PSO-AC, exploits a phase-sliding oscillation phenomenon the authors observed and validated. Offline, 22 participants produced a mean control/non-control accuracy of 94.2% from just two seconds of EEG data, a result the authors say outperforms a state-of-the-art baseline; 9-class decoding also kept its edge across different time delays. Online, in robotic-arm experiments, the method delivered more stable command triggering and higher control efficiency, with the command cost per successful trial falling from 4.30 to 1.14. The paper was published in the International Journal of Neural Systems on August 21, 2026.

Researchers proposed EFS-Net, an end-to-end network that aligns fast EEG activity with slower fNIRS haemodynamic signals through temporal, spatial and cross-modal branches. In subject-specific leave-one-session-out validation, the model reached 77.71% accuracy on a word-generation dataset and 81.69% on a mental-arithmetic dataset.

Researchers at United International University in Bangladesh proposed SSLFF, a framework that selects frequency sub-bands for each user, fuses complementary spectral information and extracts time-localized features. The paper reports statistically significant accuracy gains over conventional methods and stable performance across parameter changes, although the abstract does not provide absolute accuracy figures.

A review published on August 20, 2026 in npj Biomedical Innovations proposes brain-inspired brain-computer interfaces (BI-BCIs), a framework that unifies neuromorphic computing with BCI design to make neurotechnology lower-power, smaller and capable of closed-loop operation. The authors, from Aarhus University, Stanford University, the University of Southern Denmark, the University of Genoa, Forschungszentrum Jülich and RWTH Aachen University, say the approach could advance neuroprosthetics and neuromodulation for neurological disorders.

DRDNet separates spatial EEG features into two temporal views, models them with a bidirectional Mamba encoder and a Transformer, and then combines them through dynamic fusion and LSTM aggregation. On a public dataset, it reached 67.74% accuracy for imagined Chinese-character strokes and 62.51% for imagined pinyin vowels, outperforming seven EEG-decoding baselines.

Researchers propose MRieHy, a multi-feature Riemannian hypergraph framework for online test-time adaptation of motor imagery BCI decoding. It aligns multi-day distributions via Riemannian means of covariance matrices, builds one hypergraph with Riemannian distance and a second with cosine similarity, fuses them with adaptively learned weights, and decodes buffered online samples after Riemannian alignment. On a private four-class ECoG dataset and two public four-class EEG datasets, MRieHy shows notable gains over state-of-the-art baselines, targeting the cross-day transferability and online operation that clinical MI-BCI still lacks.

Researchers have built an AI model that combines EEG with visual gaze to score a learner's attention and cognitive load at the same time, reporting 91.8% accuracy on attention recognition and 89.4% on cognitive load, with generalization validated on public datasets. They present it as an answer to the noise sensitivity of single-signal monitoring in online learning.

WABO appointed Jianwu Dang to its Scientific Advisory Board and named him joint scientist. Dang, a researcher at the Shenzhen Institutes of Advanced Technology and a distinguished professor at Shenzhen University of Advanced Technology, will support task-conditioned intent models that combine EEG, EMG, speech, text and task context; WABO said the work is not aimed at context-free mind reading.

The preprint was posted to arXiv on August 13, 2026, proposing a hybrid deep-learning architecture that combines a convolutional neural network with a bidirectional long short-term memory network to decode motor-imagery EEG. The author notes that motor-imagery brain-computer interfaces are seen as a promising route to flexible communication between the brain and external devices, particularly for people affected by stroke or neurodegenerative disorders, but that reliable decoding remains difficult because EEG recordings carry substantial noise and relate to underlying brain activity in complex, weakly informative ways. In the proposed architecture, the CNN learns high-level spatial and temporal representations directly from raw MI-EEG recordings, while the bi-LSTM models temporal dependencies among the extracted features. The approach was evaluated on both a publicly available dataset and a privately acquired dataset collected with an EEG acquisition system, with robust performance reported on two- and three-class motor-imagery classification and promising subject-independent decoding across the methods compared. The work is a preprint and has not been peer reviewed.

A new preprint introduces BLPM, an EEG-language foundation model that reframes EEG decoding as continuous semantic embedding prediction.

Brain-computer interface decoders can guess the right label using information unrelated to the target mental state. A team at the School of Psychology, Beijing Language and Culture University (BLCU), proposes a falsifiable substitution-test standard: candidate evidence must persist in disjoint data, survive capacity-matched substitutions of physical organization or listener templates, and remain testable after target events are excluded. Across six EEG datasets (41 participants), averaging four neural-speech margin metrics brought 5-second decoding to what the authors call a leading level; in two hierarchical interfaces, parent-stream error scores kept AUCs of 0.968 and 0.965 after all target-command events were excluded. The framework offers a test for attributing evidence in neuroscience and BCI.

Researchers from Southwest State University developed an integrated brain-computer interface system supporting both synchronous and asynchronous paradigms, achieving 60-65% recognition accuracy in EEG-based robotic device control tasks.

Researchers propose ReCIL, a rehearsal-based class incremental learning method for cross-subject motor imagery classification that lets a model learn new MI classes sequentially without retraining from scratch. Using Euclidean alignment to reduce cross-subject EEG distribution shift and global-local replay to preserve earlier-task knowledge, ReCIL achieved a good balance between plasticity and stability across three public MI datasets. The authors report it as the first study of cross-subject class incremental learning for MI classification.

The study proposes BiGSTF-Net, a multimodal architecture that exploits the complementary properties of EEG and functional near-infrared spectroscopy (fNIRS) for cognitive-state decoding: heterogeneous spatio-temporal extractors capture each modality's representations, a modal residual interaction unit provides bidirectional cross-modal guidance, and a spatio-temporal gating unit fuses intra-modal features. Under cross-session evaluation on multiple BCI datasets, BiGSTF-Net consistently outperformed representative multimodal fusion baselines; ablations validated each component and visualizations matched the known neurophysiological features of the two signals.

A preprint introduces NeuroPB, a framework that scales neural decoding by transferring pretrained behavioral representations: a motion encoder is first pretrained on large-scale motor behavior data, then limited paired neural-behavior recordings align neural activity to the behavioral representation space, and a neural encoder with a lightweight motor decoder reconstructs continuous movement. The researchers report that behavioral pretraining lifted center-out trajectory decoding R² by 11% and random-target task performance by 8%, that pretraining on robotic trajectories matched macaque-trajectory performance, and that only 10% calibration was needed to match training from scratch. The study is a preprint and has not been peer-reviewed.

A preprint introduces OSPDIM, a source-free online unsupervised domain adaptation framework for class-imbalanced label shift in online EEG brain-computer interfaces. The study argues that Riemannian alignment methods such as the Riemannian centering transform handle covariate shift but implicitly assume balanced class priors, so shifting label distributions in real online use geometrically misalign the target data; OSPDIM adds a manifold-constrained bias parameter in the Riemannian tangent-space mapping, optimized by information maximization, and estimates and corrects the geometric bias online rather than relying on global batch statistics. 2D SPD matrix simulations showed OSPDIM correcting misalignment that standard centering fails on, and across multiple motor imagery datasets it significantly outperformed standard Riemannian baselines, especially online. The study is a preprint and has not been peer-reviewed.

The study presents an EEG brain-computer interface signal-processing framework that combines hybrid MothCray channel-selection optimization — fusing moth-flame and crayfish optimization — with deep learning. After notch filtering, independent component analysis (ICA) and time-window segmentation, the MothCray algorithm identifies the most informative channels and a deep neural network adapted to EEG spatio-temporal features classifies. The model reached 93.92% accuracy on BCI Competition IV dataset IIa, ahead of existing methods.

The study presents AutoMI, a framework that uses LLM multi-agents to automatically and rapidly iterate on motor imagery EEG classification models, combining a Q-learning policy with deterministic rules and integrating planning, execution and output agents with predefined tools, plus experience tracking and rollback. Models built by AutoMI reached 77.62%, 78.08% and 83.02% accuracy on the IV2a, OpenBMI and ECUST-MI datasets — up 18.42%, 9.27% and 19.25% over automated optimization algorithms.

Researchers report FRED, a principled EEG-fNIRS fusion framework for imagined handwriting decoding, posted to arXiv as a preprint on August 4, 2026 and not yet peer reviewed. Imagined handwriting offers a temporally rich paradigm for non-invasive neural decoding, and EEG-fNIRS fusion promises complementary neural information, but fusion is typically heuristic and lacks principled treatment of frequency-band redundancy. FRED builds frequency-decorrelated temporal ensembles for imagined handwriting decoding. The ensemble reaches 0.8076/0.7242/0.7492 accuracy on the public/private/overall test partitions without test-set adaptation or output constraints, and the complete system reaches 0.8498/0.7718/0.7952, ranking fourth on the private split. A modality audit finds fNIRS-only decoding at chance (0.2511 overall), while adding fNIRS to EEG changes accuracy by only +0.0025.

The study recorded hippocampal CA1 population activity in rats performing a sequential spatial decision task in a modified T-maze, dividing the decision into initiation, running and approach phases. Local theta sequences consistently over-represented the actual choice, while prospective representations driven by choice-arm place cells shifted from predicting the actual choice during running to representing potential paths more evenly at the choice point. Integrating local and prospective features improved decoding, reaching 74.4% accuracy for future choice prediction and 78.2% for upcoming trajectory decoding.

A preprint presents STEAM, a hierarchical transfer framework for EEG decoding that pairs a dual-branch spatio-temporal encoder with a shared soft mixture-of-experts (SSMoE) module to reconcile general representation learning with paradigm-specific specialization in EEG foundation models, letting complementary representations exchange information through a compact set of soft slots. The researchers report that STEAM attained the best average rank at competitive inference cost across 7 downstream datasets and 14 evaluation settings, and that its hierarchical pretraining further lifted decoding accuracy without retraining from scratch. The study is a preprint and has not been peer-reviewed.

Baylor College of Medicine researchers analyzed more than 200 hours of intracranial recordings from eight patients with treatment-resistant obsessive-compulsive disorder, five of whom also had bilateral orbitofrontal ECoG electrodes. Neural signals alone did not decode symptom severity or momentary distress above chance; adding facial and speech features raised performance to R=0.73 for severity and R=0.34 for distress. The medRxiv preprint has not been peer reviewed.

CORTIVA, a candidate-score fusion framework for EEG- and MEG-to-image retrieval, reached 73.5% Top-1 and 95.3% Top-5 accuracy across ten participants on the 200-way THINGS-EEG2 benchmark, beating the strongest reported baseline by 10.3 and 5.4 percentage points. Instead of compressing heterogeneous visual supervision into a single embedding before ranking, the authors let separate decoding routes align to different visual targets and score the same candidate pool independently, merging only the temperature-scaled score vectors, which they say preserves complementary evidence; with a modality-specific encoder the same approach reached 42.4% Top-1 on THINGS-MEG. The work is a preprint and has not been peer reviewed.

The study presents a spatial harmonization framework built on a two-layer graph convolutional network (GCN) that maps heterogeneous EEG recordings onto a unified physical electrode layout while preserving motor imagery information, addressing electrode-configuration mismatches across MI-BCI datasets. Each trial is modeled as a graph so the GCN captures spatio-temporal relations, and harmonized EEG showed lower error than spherical spline interpolation while retaining the key temporal-spectral-spatial features. Combining real and harmonized EEG lifted EEGNet accuracy from 56.57% to 66.20% and FBCNet from 61.96% to 72.54% in within-session classification on Dataset A, and supported source-only cross-dataset transfer and target-domain fine-tuning.

This preprint introduces EasyBCI, which the authors say automates BCI preprocessing across six signal types with a two-phase large language model agent. The study is a preprint and has not been peer reviewed. On EEG with a fixed linear classifier, the authors report that all five EasyBCI backbones preserve more task-relevant separability than the manual pipeline, and that the system extends to five additional modalities spanning nearly three orders of magnitude in sampling rate.