学术文献库

The possibility of simulating in detail in-vivo experiments could be highly beneficial to the neuroscientific community. It could easily allow for preliminary testing of different experimental conditions without having to be constrained by factors such as training of the subjects or resting times between experimental trials. In order to achieve this, the simulation of the environment, of the subject and of the neural system, should be as accurate as possible. Unfortunately, it is not possible to completely simulate physical systems, alongside their neural counterparts, without greatly increasing the computational cost of the simulation. For this reason, it is crucial to limit the simulation to all physical and neural areas that are involved in the experiment. We propose that using a combination of data analysis and simulated models is beneficial in determining the minimal subset of entities that have to be included in the simulation to replicate the in-vivo experiment. In particular, we focused on a pulling task performed by mice on a robotic platform before and after lesion of the central nervous system. Here, we show that, while it is possible to replicate the behaviour of the healthy mouse just by including models of the mouse forelimb, spinal cord, and recording of the rostral forelimb area (RFA), it is not possible to reproduce the behaviour of the post-stroke mouse. This can give us insights on what other elements would be needed to replicate the complete experiment.