A model of object localization during sensing: Application to the rat somatosensory whisker system

Sensory systems in mammals and humans are tuned to localize object in the environment in an accurate and behaviorally independent manner. While many engineering approaches can be used, it is less clear how nature solves the problem of object localization. In this regards, the rat somatosensory system is a well-studied experimental model in which rats actively use their whiskers to sense and navigate their surrounding for food. We investigate in this paper how the rat neocortex could obtain an object-localizing signal that is independent of the speed at which whisking occurs. Based on various detailed accumulating neurophysiological evidence in the rat barrel (whisking) cortex, we propose a preliminary mathematical model which describes the neural population activity observed over multiple stages in different cortical layers. It is suggested that interacting sensory input streams provide a speed-independent reference signal in the form of activity levels in cortical layer 5a neurons.