MR Compatible ERF-based Robotic Device for Hand Rehabilitation after Stroke

Abstract

The signal-to-noise ratio values for Sagittal T1-Weighted Localizer images during the zone 4 experiments were calculated. For the single acquisition technique, four regions were drawn: a large circular region covering most of the test object, and three smaller circular regions placed on the background air pixel. The signal to noise ratio is given by: SNR = 0.655 x (S/SDair), where S is the mean signal intensity in the large circular region, and SDair is the average of standard deviation in the three smaller regions placed over air. Table 1 shows the SNR values for three different slices. The measurements of SNR for six different voltages, gave a mean value of 88.6 (SD=1.42) for the slice 47, 90.2 (SD=1.39) for slice 58, and 88.1 (SD=1.46) for slice 75. Results show that in all cases, the loss of SNR observed was not significant regardless of the ERF being actuated in all zones where experiments were performed since 96% of the data points lie within ± 2 SD. Discussion Our novel force-feedback device designed for hand rehabilitation is unique and MR compatible combining very high computer controlled force resistance with compact geometry. Our results demonstrated that the MR environment does not affect the ERF properties. The single acquisition technique showed the ERF device had no degradation in the MR images. The MR compatible hand device may aid in the study of brain after stroke as well as the accuracy, specificity and sensitivity of measurements of motor performance when evaluating rehabilitation. Moreover, it may help in the study about the appropriate maneuvers needed for fMRI of rehabilitation in neurological disorders or stroke. Also, such rehabilitation robotic devices may help to develop a refined MR protocol for optimal neuroimaging of stroke patients, which could become a significant tool in the investigation of stroke pathophysiology.