Computational model of the dorsal horn circuitry for innocuous touch

Abstract

Providing a naturalistic tactile sensory feedback is currently one of the key challenges in neuroprosthetics. Modeling has played a fundamental role in the development of “biomimetic” stimulation protocols, determining optimal activation patterns to be elicited at the level of the nerve. Nonetheless, current technological limitations pose the challenge of estimating the influence of “errors” in matching natural activation on higher neural representations and on the elicited sensation quality. While in the past much attention has focused on the cuneate nucleus and on sensory cortices, the idea that processing starts as early as in the spinal cord dorsal horn circuitry has gained some success. Here, we propose modeling the dorsal horn light touch circuitry with a network of Izhikevich spiking point-neurons, built on the base of recent morphological and physiological analysis of interneuron populations. We propose two simple computational experiment suggesting that the dorsal horn should not be regarded as a simple relay hub. On the contrary, our model suggests that it may perform cortically modulated spatio-temporal integration of afferent signals, and sets the ground for more in-depth analysis of early somatosensory processing stages.