I am currently working with neurofeedback by fMRI in healthy volunteers to figure out if we can induce, after a one-hour neurofeedback session, improvement of motor performance.
During the neurofeedback session participants would perform motor imagery of right hand finger movement associated with neurofeedback training.
The motor imagery task would be imagining a predefined finger tapping sequence (little, middle, ring, index, ring, little, middle) at 4Hz.
Beforehand I would expect subtle motor improvement after only 1h of motor imagery training. This is especially critical in my case because both groups (neurofeedback and control) would perform motor imagery, but one group would be performing while receiving neurofeedback information in real time.
My biggest challenge so far is how to measure subtle differences in motor performance. In other words, what are the best ways to assess improvement of motor performance in my case? Which measurement would be sensitive enough to detect differences between groups?
Should I try collecting data about tapping speed and error rate (based on the predefined sequence) or should I try to investigate myoelectric response changes both before and after motor imagery? Or both?
Thank you in advance!
Dear Theo Marins,
To analysis the motor performance, in your case I recommend you apply the myoeletric analysis, which you should find interesting responses. In fact, I believe you will can find good different response for two groups (neurofeedback and control) using the two methods, however if you use the second (eletromyographic) you can explore whether there will be differences mainly between principal muscles involved in the action. Another possibility will be if you associate both, and you can produce a new study, I'm sure.
Best regards.
Hi Theo,
I would suggest you to apply a combined quantitative myoelectric and timing analysis together with fMRI in order to obtain more robust results and interpretations (ofc if you can). I would also use comparative statistical analyses across sessions and techniques to test for differences and previously established hypotheses on motor performance. This would allow you to obtain a more complete perspective of your experiment.
As Marcelo discussed above, the myoelectric analysis would offer a more realistic and valuable perspective for the medical community (as you would obtain responses directly from the muscles), whereas the timing analysis represents a more standardized technical approach as it offers parameters that are usually employed to explore and understand the basis of timing and motor performance. Moreover, the last would also allow you to correct for some confusion factors (covariates) in your experiment, such as movement amplitude or confusing factors (this is important for example, because different individual motion amplitude would have different impact in the group BOLD signal).
Our team has recently developed a 'similar' (kinematic + fMRI) approach to explore and understand the neural correlates of human gait in controls (attached) and Parkinson's disease patients (under preparation) through a rhythmic visually-guided stepping task. However, we do not consider longitudinal effects. This will not be a difficult issue in your case, as the literature provides different statistical tools to measure this. If you have any problem, I can have a look on works related with longitudinal measurements.
Kind regards
Article MRI-Compatible Device for Examining Brain Activation Related...
Article Trade‐off between frequency and precision during stepping mo...
