Towards a method to study neurorobotic control in a rat model of spinal cord injury.

Abstract

Neurorobotic control of prosthetic devices may bea viable therapeutic intervention that provides spinal cord injury patients with the ability to use the neuronal activity of populations of single neurons to control an external device (i.e.cursor on a computer screen or robotic arm). However, we are limited by our understanding of how spinal cord injury alters the ability of these neurons to convey information about the intention to move. Therefore, there is a need to develop animal models that 1) describe how population of single neurons encode information about different behavioral tasks (skilled vs.unskilled), 2) determine how this encoding is modulated by spinal cord injury and 3) perform neurorobotic control after spinal cord injury. To address the first question, we developed a rat model of spinal cord transection to evaluate the effects of the injury on the neuronal activity related to hindlimb activity. The model consists of training the rat to press a pedal with its hindlimbs. This paper describes the method that defines both the magnitude and latency of a neuron's activity in terms of its Peri-Event Histogram relative to the animal's movements during the task. The method provides a means by which changes in neural activation can be correlated with changes in behavior.