Search

Introduction/Background Neurofeedback (NF) consists on training self-regulation of brain activity by providing real-time information about the participant brain function. Few works have shown the potential of NF for stroke rehabilitation however its effectiveness has not been investigated yet. NF approaches are usually based on real-time monitoring of brain activity using a single imaging technique. Recent studies have revealed the potential of combining EEG and fMRI to achieve a more efficient and specific self-regulation, which may be critical in clinical applications. In this case report, we tested the feasibility of applying bimodal EEG-MRI NF on stroke patients. Material and method Two chronic patients affected by left hemiplegia (subcortical lesion) participated. The protocol included a calibration step (motor imagery of hemiplegic hand) and two NF sessions (5 minutes each). The experiment was run using a NF platform performing real-time EEG-fMRI processing and NF presentation ( Fig. 1 ). The NF metaphor consisted of a ball moving on a gauge proportionally to the average BOLD and EEG activity in regions of interest (ROI) identified over the ipsilesional motor cortex during calibration. Results Both patients were able to self-regulate their brain activity during the NF sessions. The EEG activity was harder to modulate than the BOLD activity ( Fig. 2 , Fig. 3 ). The correlation of the BOLD signal with the task varied depending on the ROI targeted and was particularly high for the supplementary motor area. The patients were highly motivated to engage and satisfied with the NF animation, as assessed with a qualitative questionnaire. Conclusion We performed two NF experiments involving chronic stroke patients: preliminary results showed the feasibility and the potential of applying EEG-fMRI NF for stroke rehabilitation.

In this paper, we introduce two new features for the design of electroencephalography (EEG) based Brain-Computer Interfaces (BCI): one feature based on multifractal cumulants, and one feature based on the predictive complexity of the EEG time series. The multifractal cumulants feature measures the signal regularity, while the predictive complexity measures the difficulty to predict the future of the signal based on its past, hence a degree of how complex it is. We have conducted an evaluation of the performance of these two novel features on EEG data corresponding to motor-imagery. We also compared them to the most successful features used in the BCI field, namely the Band-Power features. We evaluated these three kinds of features and their combinations on EEG signals from 13 subjects. Results obtained show that our novel features can lead to BCI designs with improved classification performance, notably when using and combining the three kinds of feature (band-power, multifractal cumulants, predictive complexity) together., Comment: Updated with more subjects. Separated out the band-power comparisons in a companion article after reviewer feedback. Source code and companion article are available at http://nicolas.brodu.numerimoire.net/en/recherche/publications

Classifying electroencephalography (EEG) signals is an important step for proceeding EEG-based brain computer interfaces (BCI). Currently, kernel based methods such as the support vector machine (SVM) are considered the state-of-the-art methods for this problem. In this paper, we apply Gaussian process (GP) classification to binary discrimination of motor imagery EEG data. Compared with the SVM, GP based methods naturally provide probability outputs for identifying a trusted prediction which can be used for post-processing in a BCI. Experimental results show that the classification methods based on a GP perform similarly to kernel logistic regression and probabilistic SVM in terms of predictive likelihood, but outperform SVM and K-nearest neighbor (KNN) in terms of 0-1 loss class prediction error.