ERD modulations during motor imageries relate to users' traits and BCI performances

Rimbert, Sébastien; Lotte, Fabien

doi:10.1109/embc48229.2022.9871411

Cited by 12 publications

(10 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Importantly, all these techniques failed to reproduce any pattern of differences between the two conditions that was replicable at the individual level (see Figure S3). However, in the same dataset, a previous work showed that the power-spectra shows differences at the group-level 29 and the grand average of the ERD/S over the cohort showed a clear desynchronization within the beta band in the contralateral sensorimotor area in the MI condition (see Figures S4 and S5) in line with previous studies [30][31][32][33] . The fact that we could find robust individual differences while discarding most data and that we failed to do so when taking the whole data into account suggests that focusing on higher-order perturbations might be useful to capture functionally-relevant processes and, in turn, to apply them to the design of BCIs.…”

Section: Discussionsupporting

confidence: 89%

Measuring Neuronal Avalanches to inform Brain-Computer Interfaces

Corsi

Sorrentino

Schwartz

et al. 2022

Preprint

View full text Add to dashboard Cite

The reconfiguration of large-scale interactions among multiple brain regions underpins complex behavior. It manifests in bursts of activations, called neuronal avalanches, which can be tracked non-invasively as they expand across the brain. Responding to a new task requires brain regions to appropriately reconfigure their interactions, which might in turn affect the path of propagation of neuronal avalanches, providing a readily-accessible read out of the processes sustaining behavior. As such, neuronal avalanche measures constitute natural candidates to design brain-computer interfaces. To test this hypothesis, we used source-reconstructed magneto/electroencephalography, comparing resting-state to motor imagery conditions during a BCI protocol. For each experimental condition, we computed an individual avalanche transition matrix, to track the probability that an avalanche would spread across any two regions. Then, we selected those edges whose transition probabilities significantly differed between conditions, at the individual level and in the majority of the participants. We found a robust topography of the edges that were affected by the execution of the task, which mainly hinge upon the premotor regions. Furthermore, we related the individual differences to the task performance, showing that significant correlations are predominantly positive and involve edges connecting pre/motor regions to parietal ones. Our results show that the pattern of propagation of large-scale perturbations are related to behavior, and that they can be used to optimize brain-computer interfaces.

show abstract

Section: Discussionsupporting

confidence: 89%

Measuring Neuronal Avalanches to inform Brain-Computer Interfaces

Corsi

Sorrentino

Schwartz

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Note that the BCI may also use other patterns than the ERD to discriminate MIs, which would explain why the intrinsic variability of the ERD or the baseline is not sufficient to explain the variability of BCI performance. The analysis of user traits (e.g., age, gender, motivation, personality) could also help to understand the variability of ERDs and BCI performances [19].…”

Section: Discussionmentioning

confidence: 99%

Is Event-Related Desynchronization variability correlated with BCI performance?

Rimbert

Trocellier

Lotte

2022

2022 IEEE International Conference on Metrology for Extended Reality, Artificial Intelligence and Neural Engineering (MetroXRAI

View full text Add to dashboard Cite

Despite current research, the relationship between the variability of Event-Related Desynchronization (ERD) generated during Motor Imagery (MI) tasks and MI-BCI performances is still not well understood. Indeed, numerous studies have previously shown that there is a lot of inter-subject and intra-subject variability in ERD patterns, but difficulties remain to understand the origin of such variability. This lack of knowledge about variability of cerebral motor patterns limits the possibilities of improving the performance of BCIs, which remains quite poor on average. We believe that a better understanding of the variability of ERDs during BCI use is crucial for developing effective interfaces. Indeed, analysis of intertrial ERDs and their variability throughout the experimental session during MI are largely neglected in most studies, which have mainly focused on identifying ERD patterns averaged across trial and possibly across participants. In this study, we propose to analyze large MI-BCI databases (n=75 subjects) and investigate how the inter/intra-individual variability of the cerebral motor patterns underlying the right-hand and lefthand MIs task (i.e., ERDs) is associated to BCI performance. Our study revealed that although ERD amplitude and baseline power are correlated with BCI performances, variability of ERD amplitude or baseline power are not.

show abstract

“…Raw EEG data was first band-pass filtered using a 3rd order, zero-phase Butterworth filter with cutoff frequencies 8 and 30 Hz, as sensorimotor rhythms have been established as being modulated during MI performance (Yger et al, 2016 ; Rimbert and Lotte, 2022 ; Shuqfa et al, 2023 ). Active MI periods (6.5-s segments from the offline, while segments of varying length from the online recording session) were isolated, and then further segmented into 1-s long epochs, simulating a sliding window analysis with a step size of 62.5 ms (i.e., 93.75% overlap among consecutive windows in a trial).…”

Section: Methodsmentioning

confidence: 99%

Combining detrended cross-correlation analysis with Riemannian geometry-based classification for improved brain-computer interface performance

Racz,

Kumar,

Kaposzta

et al. 2024

Front. Neurosci.

View full text Add to dashboard Cite

Riemannian geometry-based classification (RGBC) gained popularity in the field of brain-computer interfaces (BCIs) lately, due to its ability to deal with non-stationarities arising in electroencephalography (EEG) data. Domain adaptation, however, is most often performed on sample covariance matrices (SCMs) obtained from EEG data, and thus might not fully account for components affecting covariance estimation itself, such as regional trends. Detrended cross-correlation analysis (DCCA) can be utilized to estimate the covariance structure of such signals, yet it is computationally expensive in its original form. A recently proposed online implementation of DCCA, however, allows for its fast computation and thus makes it possible to employ DCCA in real-time applications. In this study we propose to replace the SCM with the DCCA matrix as input to RGBC and assess its effect on offline and online BCI performance. First we evaluated the proposed decoding pipeline offline on previously recorded EEG data from 18 individuals performing left and right hand motor imagery (MI), and benchmarked it against vanilla RGBC and popular MI-detection approaches. Subsequently, we recruited eight participants (with previous BCI experience) who operated an MI-based BCI (MI-BCI) online using the DCCA-enhanced Riemannian decoder. Finally, we tested the proposed method on a public, multi-class MI-BCI dataset. During offline evaluations the DCCA-based decoder consistently and significantly outperformed the other approaches. Online evaluation confirmed that the DCCA matrix could be computed in real-time even for 22-channel EEG, as well as subjects could control the MI-BCI with high command delivery (normalized Cohen's κ: 0.7409 ± 0.1515) and sample-wise MI detection (normalized Cohen's κ: 0.5200 ± 0.1610). Post-hoc analysis indicated characteristic connectivity patterns under both MI conditions, with stronger connectivity in the hemisphere contralateral to the MI task. Additionally, fractal scaling exponent of neural activity was found increased in the contralateral compared to the ipsilateral motor cortices (C4 and C3 for left and right MI, respectively) in both classes. Combining DCCA with Riemannian geometry-based decoding yields a robust and effective decoder, that not only improves upon the SCM-based approach but can also provide relevant information on the neurophysiological processes behind MI.

show abstract

ERD modulations during motor imageries relate to users' traits and BCI performances

Cited by 12 publications

References 19 publications

Measuring Neuronal Avalanches to inform Brain-Computer Interfaces

Measuring Neuronal Avalanches to inform Brain-Computer Interfaces

Is Event-Related Desynchronization variability correlated with BCI performance?

Combining detrended cross-correlation analysis with Riemannian geometry-based classification for improved brain-computer interface performance

Contact Info

Product

Resources

About