fMRI methods for proximal upper limb joint motor testing and identification of undesired mirror movement after stroke.

fMRI has been used to characterize the abnormal brain activity after stroke during attempted motor tasks, the change in brain activity accompanying spontaneous motor recovery, and response to interventions. However, many patients after stroke exhibit abnormally high effort during attempted movements, including undesired movements of the supposed quiescent, uninvolved limb, which could confound fMRI measures. We developed a method of identifying the potentially confounded scans, using EMG measures of muscle activity in the supposed quiescent limb. We found that there was no interference in the MRI signal from the EMG data acquisition system, during simultaneous use of both MRI and EMG. For EMG signal acquisition, as expected, we were able to identify EMG signal free of MRI noise contamination during the inter-scan interval between any given scan and its subsequent scan. We tested movement of the involved limb. We determined that when undesired muscle activation was present in the uninvolved, supposed quiescent limb, there was an over-estimation of the number of active voxels ranging from 10 to 11, depending upon the ROI.