Adaptive Common Spatial Pattern for single-trial EEG classification in multisubject BCI

Song, Xiaomu; Yoon, Suk-Chung; Perera, Viraga

doi:10.1109/ner.2013.6695959

Cited by 22 publications

(11 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For instance (Sun 2006) and (Shenoy 2006) proposed to re-optimize the CSP filters as a new batch of labelled data becomes available. Later, (Zhao 2008), and (Song 2013) proposed new algorithms to incrementally update the CSP spatial filters without the need to reoptimize everything. (Tomioka 2006) proposed a method to adapt spatial filters to changing EEG data class distribution.…”

Section: Adaptive Feature Extractionmentioning

confidence: 99%

A Generic Framework for Adaptive EEG-Based BCI Training and Operation

Mladenović¹,

Mattout²,

Lotte³

2018

Brain–Computer Interfaces Handbook

View full text Add to dashboard Cite

Section: Adaptive Feature Extractionmentioning

confidence: 99%

A Generic Framework for Adaptive EEG-Based BCI Training and Operation

Mladenović¹,

Mattout²,

Lotte³

2018

Brain–Computer Interfaces Handbook

View full text Add to dashboard Cite

“…Classifiers are trained offline on the sampled activity, yielding a model that maps the underlying features to a set of outputs (e.g., left vs. right imagined movement), and are then employed online to map the recorded activity to an output in real time ( Iturrate et al, 2020 ). Common machine learning algorithms used in BCI applications for decoding a subject’s imagined movement include neural networks ( Pfurtscheller and Neuper, 2001 ; Hazrati and Erfanian, 2010 ), linear discriminant analysis ( Pfurtscheller and Neuper, 2001 ; Vidaurre et al, 2011 ; Llera et al, 2014 ), and support vector machines ( Song et al, 2013 ), while during the last few years the field has experienced a steady shift of interest toward more elaborate deep learning architectures ( Roy et al, 2019 ).…”

Section: Current Approaches To Non-invasive Motor Imagery Based Brain-computer Interfacesmentioning

confidence: 99%

An Impending Paradigm Shift in Motor Imagery Based Brain-Computer Interfaces

2022

View full text Add to dashboard Cite

The development of reliable assistive devices for patients that suffer from motor impairments following central nervous system lesions remains a major challenge in the field of non-invasive Brain-Computer Interfaces (BCIs). These approaches are predominated by electroencephalography and rely on advanced signal processing and machine learning methods to extract neural correlates of motor activity. However, despite tremendous and still ongoing efforts, their value as effective clinical tools remains limited. We advocate that a rather overlooked research avenue lies in efforts to question neurophysiological markers traditionally targeted in non-invasive motor BCIs. We propose an alternative approach grounded by recent fundamental advances in non-invasive neurophysiology, specifically subject-specific feature extraction of sensorimotor bursts of activity recorded via (possibly magnetoencephalography-optimized) electroencephalography. This path holds promise in overcoming a significant proportion of existing limitations, and could foster the wider adoption of online BCIs in rehabilitation protocols.

show abstract

“…This process was repeated for successive blocks of testing data, hence generally adaptively improving the quality of the CSP and classifiers as more testing data are encountered. Along the same line of ideas, [52] also adapted the CSP spatial filters in a unsupervised way. To do so, they measured three types of differences/similarities between the (unlabeled) test trials and the training trials in order to adapt the estimated covariance matrices from each class with the test trials; with a stronger adaptation for the class for which the test trials were more similar.…”

Section: Semi-supervised Learningmentioning

confidence: 99%

“…The data sets used, the methods studied and the evaluation results are described in the following sections. Multi-users covariance matrix estimation [35], [36], [33] X Multi-task learning [38], [39], [40], [41] Domain adaptation [43], [44], [45], [42] Semi-supervised learning [49], [50], [51], [52] X A priori physiological information [53], [54], […”

Section: Evaluations and Analysesmentioning

confidence: 99%

Signal Processing Approaches to Minimize or Suppress Calibration Time in Oscillatory Activity-Based Brain–Computer Interfaces

2015

View full text Add to dashboard Cite

One of the major limitations of Brain-Computer Interfaces (BCI) is their long calibration time, which limits their use in practice, both by patients and healthy users alike. Such long calibration times are due to the large between-user variability and thus to the need to collect numerous training electroencephalography (EEG) trials for the machine learning algorithms used in BCI design. In this paper, we first survey existing approaches to reduce or suppress calibration time, these approaches being notably based on regularization, user-to-user transfer, semi-supervised learning and apriori physiological information. We then propose new tools to reduce BCI calibration time. In particular, we propose to generate artificial EEG trials from the few EEG trials initially available, in order to augment the training set size. These artificial EEG trials are obtained by relevant combinations and distortions of the original trials available. We propose 3 different methods to do so. We also propose a new, fast and simple approach to perform user-to-user transfer for BCI. Finally, we study and compare offline different approaches, both old and new ones, on the data of 50 users from 3 different BCI data sets. This enables us to identify guidelines about how to reduce or suppress calibration time for BCI.

show abstract

Adaptive Common Spatial Pattern for single-trial EEG classification in multisubject BCI

Cited by 22 publications

References 10 publications

A Generic Framework for Adaptive EEG-Based BCI Training and Operation

A Generic Framework for Adaptive EEG-Based BCI Training and Operation

An Impending Paradigm Shift in Motor Imagery Based Brain-Computer Interfaces

Signal Processing Approaches to Minimize or Suppress Calibration Time in Oscillatory Activity-Based Brain–Computer Interfaces

Contact Info

Product

Resources

About