EEG and Brain-Computer Interfaces

This paper demonstrates a brain-to-voice neuroprosthesis that synthesizes speech from neural activity, enabling real-time voice modulation with closed-loop feedback.

The paper introduces a method to localize interacting sources from EEG and MEG data by maximizing source-to-sensor coupling, which is faster and more statistically powerful than conventional approaches.

The paper shows error detection signals in speech motor cortex during neuroprosthesis use, enabling error identification with 86% accuracy and revealing distinct patterns for different error types.

This study evaluates how artifact correction and rejection affect SVM-based EEG decoding performance, finding that their combination rarely improves results but correction may still be necessary to avoid confounds.

The paper develops and validates artifact-aware sleep stage classifiers for forehead EEG, showing robust performance across multiple devices and sleep stages.

This paper introduces MIBRAIN, a framework for interindividual neural decoding using brain functional network modeling, showing improved decoding accuracy with multi-subject data integration.

This paper demonstrates that integrating large language models with neural and oculomotor signals improves BCI control accuracy for selecting arbitrary posts in a social networking service.

This paper introduces a multilayer MEG source modelling framework to enable depth-resolved laminar inference in humans, showing it aligns with known neural circuit motifs.

This paper improves multimodal EEG-fTCD brain-computer interfaces using FBCSP and envelope features, achieving high classification accuracies for motor imagery and mental rotation tasks.

The paper introduces a neurocognitive efficiency score (NES) combining reaction time and EEG data to quantify energy efficiency in cognitive tasks and demonstrates its utility across biological states.

This paper develops a calibration-free P300 BCI system with real-time performance comparable to offline benchmarks, using pre-trained models and a small electrode setup.

The paper evaluates regularized regression and clustering approaches to predict fMRI activity from EEG data, finding that clustering improves predictability but generalization remains limited.

The paper examines the temporal alignment between EEG responses and neural network activations, finding that early EEG components match low-level features and later components match semantic representations.

This paper introduces NoMAD, a method that stabilizes brain-computer interfaces by aligning latent dynamics, achieving stable decoding without supervised recalibration over long periods.

This paper introduces the SIS algorithm to improve cognitive state detection from EEG signals, achieving 29.8% better classification than traditional methods.

This paper uses SSVEP responses to a frequency-tagging flicker to detect attention lapses during prolonged vigilance tasks, finding that SSVEP outperforms other EEG markers.

The paper compares neural decoding algorithms for brain-machine interfaces to find the optimal combination of signal processing, feature extraction, and classifiers.

This paper proposes a framework for improving EEG pathology diagnosis by fusing multi-domain features and reducing redundancy, achieving better performance on a clinical dataset.

This paper proposes a meta plasticity model to explain BCI learning and skill consolidation, showing that synaptic mechanisms operate across multiple time scales to improve control proficiency.

This study identifies time-resolved MEG features that generalize across individuals for decoding visual stimuli, showing early transient features outperform EEG wavelet features and suggest shared neural coding mechanisms.

This study evaluates polarity-considered EEG microstate labeling for classifying oddball stimuli in auditory-visual tasks, showing improved classification accuracy with decision-tree models.

This study uses EEG data from a silent reading task to classify word categories using machine learning, achieving 100% accuracy with spatial dynamics and 85% with functional connectivity models.

This paper introduces a stretchable, hair-compatible EEG system that maintains signal integrity over long-term use and validates its performance in brain-computer interface tasks.

This paper presents a wireless EEG sensor network platform that enables flexible, discreet, and robust multi-channel EEG recordings with reduced motion artefacts.

This paper demonstrates that neural activity from ventral motor cortex can enable high-performance cursor control and click for brain-computer interfaces.

This paper develops a non-invasive brain-spine interface using EEG and transcutaneous spinal cord stimulation to detect movement onset and improve motor rehabilitation.

This paper introduces Meta-AlignNN, a meta-learning framework that improves BCI stability across subjects, time, and tasks by leveraging neural population dynamics consistency.

This paper uses vision transformers to distinguish between physiological ripples and epileptiform discharges in hippocampal recordings, achieving performance comparable to human experts.

The paper compares MEG and EEG for auditory attention-based brain-computer interfaces, finding that EEG with optimal channel placement can achieve comparable performance to MEG.

The paper shows scalp EEG can predict intracranial brain activity in humans, with accuracy varying by region, depth, and frequency.