Microstimulation: Principles, Techniques, and Approaches to Somatosensory Neuroprosthesis.

The power of movement of electrically charged particles has been used to alleviate an array of illnesses and help control some human body parts. Microstimulation, the electrical current-driven excitation of neural elements, is now being aimed at brain-machine interfaces (BMIs), brain-controlled external devices that improve quality of life for people such as those who have lost the ability to use their limbs. This effort is motivated by behavioral experiments that indicate a direct link between microstimulation-induced sensory experience and behavior, pointing to the possibility of optimizing and controlling the outputs of BMIs. Several laboratories have focused on using electrical stimulation to return somatosensory feedback from prosthetic limbs directly to the user's central nervous system. However, the difficulty of the problem has led to limited success thus far, and there is a need for a better understanding of the basic principles of neural microstimulation. This article provides a review of the available literature and some recent work at Downstate Medical Center and Columbia University on microstimulation of the primate and rodent somatosensory (S1) cortex and the ventral posterolateral thalamus. It is aimed at contributing to the existing knowledge base to generate good behavioral responses and effective, BMI-appropriate somatosensory feedback. In general, the threshold for the particular brain tissue in response to current-amplitude has to be determined by rigorous experimentation. For consistently reproducible results, hardware and thresholds for microstimulation have to be specified. In addition, effects on motor functions, including unwanted side effects in response to the microstimulation of brain tissue, must be examined to take the field from bench to bedside.

Tactile representation in somatosensory thalamus (VPL) and cortex (S1) of awake primate and the plasticity induced by VPL neuroprosthetic stimulation.

To further understand how tactile information is carried in somatosensory cortex (S1) and the thalamus (VPL), and how neuronal plasticity after neuroprosthetic stimulation affects sensory encoding, we chronically implanted microelectrode arrays across hand areas in both S1 and VPL, where neuronal activities were simultaneously recorded during tactile stimulation on the finger pad of awake monkeys. Tactile information encoded in the firing rate of individual units (rate coding) or in the synchrony of unit pairs (synchrony coding) was quantitatively assessed within the information theoretic-framework. We found that tactile information encoded in VPL was higher than that encoded in S1 for both rate coding and synchrony coding; rate coding carried greater information than synchrony coding for the same recording area. With the aim for neuroprosthetic stimulation, plasticity of the circuit was tested after 30 min of VPL electrical stimulation, where stimuli were delivered either randomly or contingent on the spiking of an S1 unit. We showed that neural encoding in VPL was more stable than in S1, which depends not only on the thalamic input but also on recurrent feedback. The percent change of mutual-information after stimulation was increased with closed-loop stimulation, but decreased with random stimulation. The underlying mechanisms during closed-loop stimulation might be spike-timing-dependent plasticity, while frequency-dependent synaptic plasticity might play a role in random stimulation. Our results suggest that VPL could be a promising target region for somatosensory stimulation with closed-loop brain-machine-interface applications.