Frequency-Decorrelated Temporal Ensembles for EEG--fNIRS Imagined-Handwriting Decoding

Decoding silent reading from non-invasive EEG

Non-invasive decoding of inner speech faces a fundamental data problem: a corpus pairing brain activity with a person's spontaneous inner monologue cannot be collected, and the available proxy paradigms (cued repetitive and retrospectively reported generative inner speech) are slow to acquire, poorly time-locked, and subject compliance is unverifiable. We therefore treat silent reading as a scalable proxy task and ask how much lexical and semantic information a contrastive decoder can extract from it. We report an open-vocabulary analysis of approximately 240,000 word presentations recorded from a single densely-sampled participant across 393 runs (ca. 49 h) of 19-channel dry-electrode EEG. Words from continuous narrative text were presented in rapid serial visual presentation, with typography randomised on every trial to partially decorrelate word identity from low-level visual form. A convolutional EEG encoder, optionally followed by a causal transformer, was trained with a CLIP-style contrastive objective to align short EEG windows with hidden-state embeddings of the presented word taken from a large language model. Decoding, evaluated as word-grouped top-10 retrieval against permutation baselines, was reliably above chance, extended to mid-frequency and rare words, and scaled log-linearly with training-data volume with no sign of saturation. Removing occipital and posterior-temporal electrodes reduced the word-level gain by roughly one third but left context tracking unchanged. Control analyses separate word-level decoding from narrative context tracking and from a non-neural positional prior introduced by the transformer's positional embedding. These results establish that open-vocabulary word-level information is recoverable from EEG during silent reading, and that decoding is data-limited rather than saturated.

EEG-PRIME: Prototype-Aligned Representation Learning with Multi-Level Conditioning for EEG Decoding

Electroencephalography (EEG) decoding models often generalize poorly across datasets and subjects due to domain shifts in acquisition protocols and individual neurophysiology. We propose EEG-PRIME, a two-stage EEG foundation model for cross-dataset multi-task decoding. EEG-PRIME combines masked pretraining with prototype-aligned instruction tuning to enable instruction-aware and subject-invariant decoding across diverse BCI paradigms. During pretraining, an EEG encoder learns transferable representations through masked reconstruction with frequency-cutoff spectral augmentation. During instruction tuning, EEG-PRIME incorporates task-semantic, dataset-specific, and subject-invariant conditioning. The resulting conditioning signal modulates the Q-Former through Layer-wise Query Modulation, while frozen text embeddings of class labels serve as prototypes for cosine-similarity-based prediction across heterogeneous label spaces. Experiments on sixteen datasets covering motor imagery, emotion recognition, ADHD detection, covert speech, and mental workload show consistent improvements over state-of-the-art baselines and prior EEG foundation models under cross-subject settings. On two additional held-out datasets, EEG-PRIME achieves balanced accuracy comparable to within-session calibration models without target-domain optimization, calibration, or linear probing, demonstrating promising zero-shot transfer capability.

Multi-Domain Feature Fusion and Channel Attention in an Inception-Based Architecture for Motor Imagery EEG Decoding

Background/Objectives: Existing motor imagery (MI) EEG decoding is often limited by small datasets, affecting generalization reliability. This study aims to robustly decode multi-limb MI intentions. Methods: We collected an MI-EEG dataset from 292 participants (242 young adults, 50 older adults) performing left/right-arm and left/right-leg imagery. After extracting time-, frequency-, and channel correlation features, we proposed an SE-EEG-Inception model for classification. Results: Evaluated under a strict intra-subject cross-validation protocol, the model achieved a mean 4-class accuracy of 89.4%. For binary tasks, accuracies reached 88.3% (left vs. right arm) and 90.0% (left vs. right leg). Conclusions: The model successfully distinguishes predictive EEG features across different and symmetric limbs. Crucially, this high classification performance demonstrates data-driven predictive utility rather than mechanistic proof of neural differences, providing an offline proof-of-concept for multi-limb BCI control.

Cross-variability decoding for motor imagery EEG signals: a comprehensive review

Motor Imagery-based (MI) Electroencephalography (EEG) has emerged as a leading solution in non-invasive Brain-Computer Interface (BCI) systems, leveraging its strong motor intention correlation to enable reliable neural decoding. However, practical implementation of MI confronts three persistent challenges: low signal-to-noise ratio, substantial variability across subjects or over time, and inherent signal nonstationarity. These fundamental limitations continue to hinder the widespread adoption and operational reliability of MI BCI systems. Despite advances in cross-variability decoding methods, there is a lack of systematic syntheses to guide technological evolution in MI BCI. To address these challenges, this review presents a comprehensive taxonomy of MI EEG cross-variability decoding studies from 2020 to 2025, systematically organizing advances in deep learning and transfer learning. We critically evaluate core algorithmic approaches, including Convolutional Neural Networks (CNN), transformers, feature alignment, domain adaptation, and meta-learning. We then explore the underlying mechanisms of these methods and assess their efficacy across key variability paradigms (mainly cross-subject and cross-session scenarios). Finally, we summarize key findings, highlight unresolved challenges, and outline promising future research directions. These advancements hold significant potential to bridge the gap between laboratory-based MI and real-world clinical and consumer applications.

DRDNet: a dual-view representation decoupling network for handwriting imagery EEG classification

Objective. Handwriting imagery (HI) based on electroencephalography (EEG) offers a non-invasive route to text input and complex intention expression for brain-computer interfaces (BCIs). However, HI EEG decoding is challenged by low signal-to-noise ratios, non-stationarity, cross-session distribution shifts, and the coexistence of continuous temporal trends and local high-response patterns. Approach. We propose a Dual-View Representation Decoupling Network (DRDNet) for HI EEG classification under within-subject cross-session and cross-subject evaluation settings. DRDNet first constructs two complementary temporal views from spatial EEG features using average and max pooling, corresponding to smooth trend-oriented and salient response-oriented representations. These views are then modeled by a bidirectional Mamba encoder and a Transformer encoder, respectively, and adaptively integrated through a time-step-level dynamic fusion mechanism followed by long short-term memory-based temporal aggregation. The method is evaluated on two tasks from a public HI EEG dataset: Chinese character stroke HI (CCSHI) and pinyin single-vowel HI (SVHI). Main results. DRDNet achieved average accuracies of 67.74% and 62.51%, with Cohen’s kappa scores of 0.5968 and 0.5502, on CCSHI and SVHI, respectively. Under the cross-subject setting, DRDNet also achieved the best average accuracies of 62.54% and 54.40% on CCSHI and SVHI, respectively. DRDNet outperformed seven representative EEG decoding baselines under both evaluation settings. Confusion matrices, feature visualization, ablation studies, structural variant analysis, and complexity evaluation further showed improved feature separability and a favorable balance between decoding performance and computational efficiency. Significance. The results indicate that decoupling HI EEG into complementary temporal views and matching them with heterogeneous temporal encoders provides an effective representation learning strategy for robust non-invasive handwriting BCI decoding.

A balanced multimodal decoding framework of EEG and fNIRS for motor imagery