EEG and Brain-Computer Interfaces

This paper introduces an optimized deep learning model for real-time classification of student confusion using EEG data, achieving higher accuracy than existing methods.

The paper develops a method to select optimal filter settings for ERP data by maximizing signal-to-noise ratio while minimizing waveform distortion.

The paper recommends optimal filter cutoffs for seven common ERP components to balance signal-to-noise ratio and waveform distortion across different noise levels.

The paper identifies temporal correlations between natural single-shot videos and EEG signals, showing that object-based features outperform traditional video features in capturing these correlations.

This study investigates biases in EEG-based spatial auditory attention decoding, finding that simple models can overfit to eye-gaze and trial-specific confounds, leading to inflated accuracy.

This paper shows that recurrent neural networks (RNNs) can be adapted for real-time neuroprosthetic control by modifying their training data's temporal structure, enabling better performance than standard linear methods.

The paper investigates beta burst activity in motor imagery tasks, showing that burst-based features can match or exceed traditional beta power methods in decoding brain signals.

This paper evaluates artifact correction and rejection methods in ERP research, finding that correction reduces confounds and rejection reduces noise, improving data quality.

This paper improves motor imagery-based BCI by analyzing beta burst waveforms instead of beta power, showing better classification performance and information transfer rates.

This paper develops a method for speech recognition from MEG data using covariance filters, achieving reliable recognition with minimal errors in 7 subjects.

This study compares KalmanNet, an explainable BMI decoder, with deep-learning and Kalman filter models, finding it achieves comparable performance while offering interpretability.

This paper develops a method to decode internal decision making during reverse engineering tasks using neural markers and user behavior metrics.

This paper introduces a method to detect trial-wise cognitive events in neural time-series, showing it can estimate event patterns, sensor contributions, and inter-event intervals from EEG/MEG data.

This study predicts age from at-home EEG data during sleep and meditation, achieving R2 scores of 0.3-0.5, comparable to lab-grade EEG benchmarks.

The paper introduces the GLM-Spectrum framework for analyzing power spectra with confound and covariate modeling, showing its ability to quantify condition differences and age effects in EEG data.

This study uses periliminal and subliminal flickers to measure spatial attention via SSVEP, showing improved user experience and classification accuracy.

This study introduces two novel convolutional neural networks for MDD diagnosis from EEG data, finding higher beta and delta power and right hemisphere correlation in MDD patients.

The paper develops a supervised spatial filter to remove speech artifacts from intracranial EEG recordings while preserving neural activity.

This paper develops automated methods using ICA and HMM to detect and localize interictal epileptiform discharges in MEG data, showing high sensitivity and potential to improve clinical practice.

The paper addresses two common issues in synchronized multimodal EEG recordings: jitter and latency, using a Lab Streaming Layer-based system for precise time alignment.

This study uses neuronal avalanches to improve brain-computer interfaces by showing their task-specific differences can be decoded better than current methods.

This study introduces a single-channel EEG artifact detection method using spectral slope, achieving high accuracy in identifying contaminated epochs in rat recordings.

This study improves major depressive disorder diagnosis using raw multichannel EEG with transfer learning from single-channel sleep stage data, achieving better performance than previous methods.

This study develops a robust and explainable deep learning model for schizophrenia diagnosis using EEG data, showing improved performance and prioritization of specific brain regions and frequency bands.

This meta-analysis evaluates neurofeedback training's impact on attention in healthy adults, finding a significant overall effect but no difference from sham training.

This study demonstrates a closed-loop ultrasonic brain-machine interface that enables nonhuman primates to control movements using real-time neurovascular activity predictions.

This paper improves M/EEG-based regression models by applying covariance-based domain adaptation techniques to address dataset shifts across different recording contexts.

This study improves image reconstruction from brain activity by incorporating semantic information, enabling reconstruction of both seen and imagined images.

This paper introduces Burst c-VEP, a novel visual stimulus design for BCI that improves classification accuracy and user experience with minimal calibration data.

This paper introduces an inter-brain attention coupling method to detect learning-related attention in primary school students and finds it correlates with academic performance.