Mazorow Dissertation Abstract
Enhancing reach-to-grasp movements in two and three dimensions using supplemental vibrotactile kinesthetic feedback: information encoding, extended training, and effects of sensorimotor and neuropsychological function
Dissertation Date: April 7, 2026
Approximately half of stroke survivors experience decreased sensations of limb position and movement, leading to deficits in upper extremity motor control. Although visual feedback of the arm and hand can partly compensate, this reliance results in slow, jerky movements and reduction in awareness of the surrounding environment. An emerging body of literature suggests an alternate compensatory approach: providing supplemental vibrotactile feedback to restore closed-loop control of the arm. Previous research found that healthy individuals and some survivors of stroke can use vibrotactile feedback to enhance the accuracy of 2-dimensional (2D) reaching tasks performed without visual feedback when the vibrotactile feedback was applied to the non-moving limb. The dissertation explores: (1) how vibrotactile feedback may best encode information about the state of the moving limb; (2) the effects of extended 3-dimensional (3D) vibrotactile training on stroke survivors’ reach-to-grasp performance; and (3) the relationship between 3D vibrotactile skill acquisition and clinical assessments of sensorimotor and neuropsychological function.
First, we compared the utility and subjective user experience for two biologically inspired ways to encode movement-related information into 2D vibrotactile feedback. One encoding scheme mimicked visual feedback encoding by converting real-time hand position in a Cartesian frame of reference into vibrotactile kinesthetic feedback. The other approach mimicked proprioceptive encoding by encoding real-time arm joint angle information. After a brief training period without concurrent visual feedback, 15 healthy young adults were able to use both forms of supplemental feedback to improve reach accuracy over performance using proprioception alone. Cartesian encoding yielded greater improvements in reach accuracy. Participant responses on user experience surveys indicated that both encoding schemes were motivating and both yielded passable user satisfaction scores. However, only Cartesian encoding had passable usability. This informed our decision to use Cartesian endpoint encoding in future studies.
Second, we redesigned the vibrotactile interface to provide feedback for 3D movements. Fifteen healthy young adults underwent three days (~2 hours total) of reach-to-grasp training of movements in 3D space. Experimental conditions promoted learning a mapping from 3D hand position to patterns of vibrotactile feedback. This cohort utilized 3D vibrotactile feedback without concurrent visual feedback to achieve greater accuracy than when they had to rely on proprioception only. Next, seven survivors of stroke underwent 18 days of training (9 hours total). When reaching with the supplemental feedback after training, three participants improved reach accuracy, two improved temporal efficiency, and three improved spatial efficiency (although one worsened with regard to spatial efficiency). Only one participant experienced performance improvements with vibrotactile feedback without concomitant improvements when reaching with proprioception alone. These results suggest that practical benefits after stroke may accrue more from hours of focused reach training than from the additional sensory information provided by the vibrotactile interface.
Finally, we identified sensorimotor and neuropsychological factors that impact a person's ability to integrate supplemental vibrotactile kinesthetic feedback into the ongoing control of goal-directed reaching movements. Seven stroke survivors and eight age-matched controls underwent clinical assessments of sensorimotor and neuropsychological function before engaging in 9 hours of training with 3D vibrotactile feedback. Correlation analyses found that for stroke survivors, successful integration of vibrotactile feedback into the ongoing control of movement is constrained by motor impairment. For all participants, successful integration was constrained by psychomotor processing speed.
This dissertation provides insight into whether vibrotactile feedback is a viable long-term solution for kinesthetic augmentation. Young, healthy adults were able to use vibrotactile kinesthetic feedback to yield greater reach accuracy than when forced to reach with intrinsic proprioception alone. However, results were mixed with older healthy adults and survivors of stroke, suggesting that some, but not all, potential users may benefit from the proposed compensatory technology.
