Validation of bimanual-coordinated training supported by a new upper-limb rehabilitation robot: a near-infrared...

Our previous work proposed a rehabilitation robot to support bimanual-coordinated training not only in active-assisted and passive-driven modes but also in active-resisted mode. However, assessment of training effect was only focused on the improvements of subjects' motion-tracking precisions. This paper presents an evaluation strategy based on variations in both cerebral activation level (CAL) and motion-tracking precision.

Fourteen healthy subjects participated in motion-tracking training in bimanual active-assisted and active-resisted modes, and in single right-limb and left-limb modes, with haemoglobin concentration and motion-tracking errors being measured simultaneously. Analyses of variance (ANOVA) of the CAL and motion-tracking errors were performed to investigate bimanual training effect and the difference between bimanual and single-limb trainings in activating the brain.

In the bimanual modes, both the CAL and motion-tracking precision significantly increased after training. And the CAL induced in the bimanual trainings were significantly greater than in the single-limb trainings.

Significant enhancement of the CAL and motion-tracking precision confirmed a positive training effect on enhancing the bimanual-coordination capability of healthy subjects. Compared to the single-limb modes, the higher CAL in the bimanual modes demonstrated the potential of the proposed bimanual training for improving the functional integrity of the two hemispheres.