Functional MRI-based identification of brain areas involved in motor imagery for implantable brain-computer interfaces

D. Hermes, M. J. Vansteensel, A. M. Albers, M. G. Bleichner, M. R. Benedictus, C. Mendez Orellana, E. J. Aarnoutse, N. F. Ramsey

Research output: Contribution to journalArticlepeer-review

39 Scopus citations

Abstract

For the development of minimally invasive brain-computer interfaces (BCIs), it is important to accurately localize the area of implantation. Using fMRI, we investigated which brain areas are involved in motor imagery. Twelve healthy subjects performed a motor execution and imagery task during separate fMRI and EEG measurements. fMRI results showed that during imagery, premotor and parietal areas were most robustly activated in individual subjects, but surprisingly, no activation was found in the primary motor cortex. EEG results showed that spectral power decreases in contralateral sensorimotor rhythms (8-24 Hz) during both movement and imagery. To further verify the involvement of the motor imagery areas found with fMRI, one epilepsy patient performed the same task during both fMRI and ECoG recordings. Significant ECoG low (8-24 Hz) and high (65-95 Hz) frequency power changes were observed selectively on premotor cortex and these co-localized with fMRI. During a subsequent BCI task, excellent performance (91%) was obtained based on ECoG power changes from the localized premotor area. These results indicate that other areas than the primary motor area may be more reliably activated during motor imagery. Specifically, the premotor cortex may be a better area to implant an invasive BCI.

Original languageEnglish (US)
Article number025007
JournalJournal of neural engineering
Volume8
Issue number2
DOIs
StatePublished - Apr 2011

ASJC Scopus subject areas

  • Biomedical Engineering
  • Cellular and Molecular Neuroscience

Fingerprint Dive into the research topics of 'Functional MRI-based identification of brain areas involved in motor imagery for implantable brain-computer interfaces'. Together they form a unique fingerprint.

Cite this