Abstract
Compared to functional magnetic resonance imaging (fMRI), functional near infrared spectroscopy (fNIRS) has several advantages that make it particularly interesting for neurofeedback (NFB). A pre-requisite for NFB applications is that with fNIRS, signals from the brain region of interest can be measured. This study focused on the supplementary motor area (SMA). Healthy older participants (N = 16) completed separate continuous-wave (CW-) fNIRS and (f)MRI sessions. Data were collected for executed and imagined hand movements (motor imagery, MI), and for MI of whole body movements. Individual anatomical data were used to (i) define the regions of interest for fMRI analysis, to (ii) extract the fMRI BOLD response from the cortical regions corresponding to the fNIRS channels, and (iii) to select fNIRS channels. Concentration changes in oxygenated (Delta [HbO]) and deoxygenated (Delta [HbR]) hemoglobin were considered in the analyses. Results revealed subtle differences between the different MI tasks, indicating that for whole body MI movements as well as for MI of hand movements Delta [HbR] is the more specific signal. Selection of the fNIRS channel set based on individual anatomy did not improve the results. Overall, the study indicates that in terms of spatial specificity and task sensitivity SMA activation can be reliably measured with CW-fNIRS.
Highlights
Of supplementary motor area activation during motor execution and motor imagery Franziska Klein1,2*, Stefan Debener[2], Karsten Witt3 & Cornelia Kranczioch[1,2]
Out of 34 participants (17 females, males) a total of participants were excluded because of the following reasons: mild cognitive decline as detected by the MoCa (2), poor functional near infrared spectroscopy (fNIRS) signal quality (2), falling asleep in the MR scanner (1), typing with the wrong hand (2), forgetting repeatedly not to execute the movements during motor imagery (MI) (1), cancelling the second session (1), insufficient cap placement (2), a beta mask that did not cover the whole cortex (1), and electromyography (EMG) signal saturation (2) and excessive movement during MI (4). The latter was the case if for a participant more than half of the trials of a single MI task were classified as movement trials in either of both sessions (fNIRS or functional magnetic resonance imaging (fMRI); for more details cf. section Electromyography (EMG))
Fig. 1A), for both motor execution (ME) LEFT and ME RIGHT a clear M1 lateralisation with comparable spatial patterns can be seen for fMRI PEAK and fMRI CHANLOCS data
Summary
Of supplementary motor area activation during motor execution and motor imagery Franziska Klein1,2*, Stefan Debener[2], Karsten Witt3 & Cornelia Kranczioch[1,2]. Functional near infrared spectroscopy (fNIRS) has gained considerable popularity over the past decades This popularity is to a large part owed to the fact that compared to the gold standard for in vivo imaging of the brain, functional magnetic resonance imaging (fMRI), less restrictions and safety concerns apply to fNIRS. FNIRS does not capture subcortical activation and is limited to superficial cortical brain r egions[4] Another limitation is that optode placement and data processing pipelines have to cope with a lack of anatomical information[4]. Software has been developed to address this issue (e.g., fOLD5 and AtlasViewer6) by guiding optode placement using anatomical information taken from standard brains While this approach is without doubt very useful, the selection of channels for subsequent analysis may benefit from the consideration of individual anatomical information. A third limitation is the relatively low fNIRS signal-to-noise ratio, largely resulting from contamination by systemic noise
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
