How does the cerebellum automate and coordinate unconscious motor sequences?
Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
In most vertebrates, the cerebellum occupies 20–30% of the brain. It has been heavily researched and there is a rich literature. Despite that, we don’t know (1) how it codes or processes information, or (2) how it is connected around the body and controls movement. We try to answer those questions for the cerebellum, in turn. The cerebellum is formed of repeating cell networks. In the first step, we model the network computation, about 30 million cells, in detail that includes individual anatomical variability. A central proposal is that network computations are the passive, unlearned result of anatomy, counter to 50 years of the idea that computations depend on learned synaptic changes. In the second step, we propose, in principle, how input to a locomotor network may be topographically organised and where output is sent, and how output is coordinated with the output of other networks. We use a model of eel-like swimming to illustrate unconscious control and propagation of skilled motor sequences. Undulating swimming was probably the original vertebrate solution to changing locations. The main structures and organisation of the vertebrate brain are highly conserved across species, suggesting that the original wiring plan of motor control may have been adapted later to other vertebrate phenotypes.
Lay summary
It is poorly understood – i.e., not understood – how the cerebellum represents information and uses it to generate motor outputs, and how outputs are coordinated with each other and with behaviour to control movement. We try to answer those questions.
The paper is in two main parts. In the first, we model the cerebellar network computation. We propose the network computation is unlearned, simply a passive effect of cell morphologies and network geometry. This is still sufficient to distil a single output rate – motor output of the cerebellum – from 80,000 input signals. The form of the computation means it is unnecessary to simplify the network to eliminate noise. That makes it possible to model the network with close attention to detail.
In the second part, we propose that the topography of input and output of cerebellar motor circuits could automate unconscious control of motor sequences. To illustrate how that could work, we model swimming. We propose that the cerebellum connects proprioception-motor loops so that movement itself can drive coordinated motor output with only sparse (go, no go) executive input. The role of the cerebellum and functional topography of connections may remain today a variation of the same principles.
