Human psychophysics of direct cortical stimulation of somatosensory cortex

Abstract

Feedback is a vital part of any control system, and somatosensory feedback is essential for efficient and precise movement. Thus, future rehabilitative brain-computer interfaces (BCIs), such as neuroprostheses, require the integration of somatosensory feedback for improved control. Such bidirectional BCIs (BBCIs) will both record from and stimulate the nervous system to restore function. Direct cortical stimulation (DCS) of the primary somatosensory cortex (S1) through electrocorticography (ECoG) electrodes is one possible method of providing sensory feedback to users. The future success of ECoG sensory stimulation will depend on our ability to elicit a wide range of useful tactile perceptions through stimulation. However, the field’s understanding of the relationship between DCS parameters, a subject’s ability to perceive DCS, and the percepts a subject experiences is limited. I have endeavored to address these limitations using DCS via macro-ECoG electrodes in human subjects who can describe their perception and experience of DCS. We first examined the psychophysics of DCS of S1, using bipolar, biphasic DCS waveforms characterized by their current amplitude, pulse phase width, pulse frequency, and stimulation train duration. We found that although perceptual thresholds vary between subjects, certain relationships between the stimulation parameters and their perceptual thresholds hold, including a previously reported finding in non-human primates and rats that perceptual charge thresholds decrease with decreasing pulse width and increasing pulse frequency. Next we assessed subjects’ response times to S1 DCS compared to their response times to haptic stimuli and found that subjects respond significantly slower to S1 DCS than to haptic touch. Finally we considered how S1 DCS could perform in a future BBCI by evaluating how S1 DCS would affect subjects’ sense of ownership over an artificial limb and exploring how subjects could use S1 DCS as feedback in a motor-based task. We found that S1 DCS could indeed evoke a sense of ownership and be used as task feedback, but may not have been a trivial feedback signal to learn. These findings have implications in the development of somatosensory feedback for BBCIs. Given that we observed similar trends in subject perception as is noted in studies using other electrical stimulation modalities, it is possible that we can apply what we have learned through these macro-ECoG DCS studies to other stimulation modalities.