Back

Development of adaptive motor control for tactile navigation

Azarfar, A.; Celikel, T.

2019-09-08 neuroscience
10.1101/762443 bioRxiv
Show abstract

Navigation is a result of complex sensorimotor computation which requires integration of sensory information in allocentric and egocentric coordinates as the brain computes a motor plan to drive navigation. In this active sensing process, motor commands are adaptively regulated based on prior sensory information. In the darkness, rodents commonly rely on their tactile senses, in particular to their whiskers, to gather the necessary sensory information and instruct navigation. Previous research has shown that rodents can process whisker input to guide mobility even prior to whisking onset by the end of the second postnatal week, however, when and how adaptive sensorimotor control of whisker position matures is still not known. Here, we addressed this question in rats longitudinally as animals searched for a stationary target in darkness. The results showed that juvenile rats perform object localization by controlling their body, but not whisker position, based on the expected location of the target. Adaptive, closed-loop, control of whisker position matures only after the third postnatal week. Computational modeling of the active whisking showed the emergence of the closed-loop control of whisker position and reactive retraction, i.e. whisker retraction that ensures the constancy of duration of tactile sampling, facilitate the maturation of sensorimotor exploration strategies during active sensing. These results argue that adaptive motor control of body and whiskers develop sequentially, and sensorimotor control of whisker position emerges later in postnatal development upon the maturation of intracortical sensorimotor circuits.

Matching journals

The top 2 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.