In memory of Cal Shearer
21st November 2024
4th Oct 2016
A recent article published in Nature Neuroscience has shown evidence of the mouse brain acting like GPS navigator in the absence of sound cue signaling the location of the reward.
Dynamic Bayesian inference allows the system to infer the environmental state under limited sensory observation conditions. The researchers used a goal-reaching task to find the two fundamental features of dynamic Bayesian inference present. These being the prediction of hidden states by the posterior parietal cortex and the updating of sensory input by the adjacent posteromedial cortex.
As the mice approached the reward site, anticipatory licking increased, even when sound cues were intermittently presented. When the mice were injected with a drug to suppress neural activity, anticipatory licking was not displayed. This suggests the parietal cortex plays a role in goal prediction.
The results of this study show how cerebral cortex realises mental stimulation using an action-dependent dynamic model.
To read the full article, please visit Nature website.