Introduction: The restoration of function through brain computer interface (BCI) will require responsive somatosensory feedback for both closed-loop motor-sensory integration, and for restoration of sensory information from the body. Work in closed-loop BCI for non-human primates has shown improved motor outcomes from sensory feedback, and recent work in human somatosensory stimulation, yields consistent sensory feedback. Continuing with this work, our group evaluated frequency as a variable for delivering different somatosensory precepts.
Methods: Two epilepsy patients, a right-handed female and male, were implanted with a subdural electrocorticography grid over the hand region of somatosensory cortex (left-sided and right-sided, respectively). Each performed a frequency discrimination target acquisition task consisting of the subject exploring two targets with their hand contralateral to the implant, without physically touching the targets. Over each target, the primary somatosensory cortex was stimulated randomly at either 50 or 100 Hz, with stable current, pulse width, and pulse duration. The subject had to determine which target had the higher frequency.
Results: The ventral surface of the tip of digit two was chosen for testing for the female subject, and the medial surface of the palm and digit 5 were chosen for the male subject. Each subject performed 25 trials. Accuracy was 98% among the two subjects. The difference in sensation was described as “more intense” and “faster” by the female subject, and “stronger” and “faster buzzing” by the male subject. No adverse events occurred.
Conclusions: The high degree of accuracy in discriminating between two frequencies while holding other variables of stimulation constant makes this a usable component to manipulate in future somatosensory BCI systems. Future work will require elucidation of more detailed discrimination among frequencies to establish just noticeable differences and the upper and lower limits of functional adjustment.
Patient Care: Providing sensory feedback in brain computer interface systems will improve the control, safety and utility of such systems. This could potentially benefit patients with decreased peripheral nerve function, due to stroke, injury or neuropathy.
Learning Objectives: 1) Describe varying somatosensory stimulation frequency as a method to provide sensory feedback in BCI systems. 2) Describe an experimental design to evaluate efficacy of changing somatosensory stimulation parameters to provide BCI feedback.