Elsevier

NeuroImage

Volume 31, Issue 1, 15 May 2006, Pages 153-159
NeuroImage

Mu rhythm (de)synchronization and EEG single-trial classification of different motor imagery tasks

https://doi.org/10.1016/j.neuroimage.2005.12.003Get rights and content

Abstract

We studied the reactivity of EEG rhythms (mu rhythms) in association with the imagination of right hand, left hand, foot, and tongue movement with 60 EEG electrodes in nine able-bodied subjects. During hand motor imagery, the hand mu rhythm blocked or desynchronized in all subjects, whereas an enhancement of the hand area mu rhythm was observed during foot or tongue motor imagery in the majority of the subjects. The frequency of the most reactive components was 11.7 Hz ± 0.4 (mean ± SD). While the desynchronized components were broad banded and centered at 10.9 Hz ± 0.9, the synchronized components were narrow banded and displayed higher frequencies at 12.0 Hz ± 1.0. The discrimination between the four motor imagery tasks based on classification of single EEG trials improved when, in addition to event-related desynchronization (ERD), event-related synchronization (ERS) patterns were induced in at least one or two tasks. This implies that such EEG phenomena may be utilized in a multi-class brain–computer interface (BCI) operated simply by motor imagery.

Introduction

A fundamental property of a neural network is the ability of neurons to work in synchrony and to generate oscillatory activity (Lopes da Silva, 1991). One prominent group of such brain oscillations has frequencies between 9–13 Hz in man and 12–15 Hz in cat and originates in sensorimotor areas. These activities are known as “rolandic mu rhythms” or “wicket rhythms” in man (Niedermeyer, 1993, Gastaut, 1952) and sensorimotor rhythms (SMRs) in cat (Chase and Harper, 1971, Howe and Sterman, 1972).

It is well known that planning and execution of hand and/or finger movement block or desynchronize the mu rhythm (Chatrian et al., 1959), and inhibition of motor behavior synchronizes the SMR (Howe and Sterman, 1972). The importance of such an enhancement of 12- to 15-Hz oscillations for biofeedback therapy was documented already in the seventies by Sterman et al. (1974) and confirmed by Egner and Gruzelier (2001) and others. It was already demonstrated that externally paced foot and tongue movement and imagination of foot movement (Pfurtscheller and Neuper, 1994, Pfurtscheller and Neuper, 1997) can enhance the hand area mu rhythm, similar as observed during reading of words (Pfurtscheller, 1992), pattern vision (Koshino and Niedermeyer, 1975) or flicker stimulation (Brechet and Lecasble, 1965). This ability to suppress or enhance the amplitude of the hand area mu rhythm consciously by directing attention to different body parts or limbs is not only of interest to suppress epileptic seizures by neurofeedback therapy (Sterman et al., 1974) but also for realizing an EEG-based brain–computer interface (BCI) with motor imagery as a mental strategy (Wolpaw et al., 2002, Pfurtscheller and Neuper, 2001).

The goals of this paper are

  • (i)

    to study the inter- and intrasubject variability of event-related EEG (de)synchronization patterns (ERD/ERS) in four motor imagery tasks,

  • (ii)

    to study whether the same or different frequency components are involved in desynchronization and synchronization patterns recorded from the same cortical areas,

  • (iii)

    to report on the distinctiveness between four different motor imagery tasks when single trials are analyzed and classified, and

  • (iv)

    to provide recommendations for the realization of a multi-class BCI with improved classification accuracy.

Section snippets

Subjects and experimental paradigm

Six female and three male healthy right-handed subjects (mean age 26.2 years, range 21–31 years) participated in this study. They sat in a comfortable armchair in an electrically shielded cabin watching a 15″ monitor from a distance of about 2 m. Each trial started with a blank screen at second 0. At second 2, a fixation cross was presented at the center of the monitor until the end of the trial at second 7. Simultaneously, a short warning tone occurred at second 2. At second 3, an arrow,

Averaged band power in alpha (mu) frequency bands during motor imagery tasks

Examples of time–frequency ERD/ERS maps from one subject are presented in Fig. 2a. The maps for electrode positions C3, Cz, and C4 show characteristic patterns of mu and beta ERD only for right and left hand motor imagery, namely, a broad-banded ERD in the 10-Hz and 20-Hz bands at electrode positions C3 and C4 with a contralateral dominance during right hand imagery. Quite different patterns are found with foot and tongue motor imagery. In the first case, an ERS in the 15-Hz band is dominant at

“Focal ERD/surround ERS” induced by motor imagery

Basically, hand motor imagery activates neural networks in the cortical hand representation area which is manifested as blocking or desynchronization of the hand area mu rhythm (mu ERD). Such a mu ERD was found in all subjects during right and left hand motor imagery with a clear contralateral dominance. Less clear is the activation of the foot representation area during foot motor imagery because of its location in the mesial wall. In this case, a midcentral mu ERD was found not in all but in

Conclusion

During performance of different motor imagery tasks, there exists not only a great intersubject variability but also a considerable intrasubject variability concerning the reactivity of upper mu components. Different types of band power changes (enhancement vs. suppression) during different imagery tasks are a prerequisite for an optimal distinctiveness between different motor imagery tasks when single trials are analyzed. We should emphasize that the ERD/ERS changes reported here were elicited

Acknowledgments

The work was funded by the European PRESENCIA project (IST-2001-37927) and the Austrian FWF project P16326-BO2. We would like to thank C. Keinrath for the data recording and B. Graimann for the support in signal processing.

References (36)

Cited by (1284)

View all citing articles on Scopus
View full text