Biomedical Engineering Reference
In-Depth Information
C H A P T E R 2
Foundations of Neuronal
Representations
additional Contributors: Sung-Phil Kim, yiwen wang
The development of BMIs relies upon the ability to decipher neuronal representations expressed by
individual neuronal elements operating in distributed ensembles, and communicating using con-
vergent and divergent principles.
1
Understanding the properties of information transmission and
determining how this information translates into commands of the motor system as a whole is one
of the cornerstones of BMI development. The intricate computational neurobiology of motor BMIs
can be elucidated by first specifying the general principles of neural representation, transmission,
and signal processing in the system, which are summarized in the following questions:
•
•
•
•
What are the functional computational elements?
What are the relevant mechanisms of information transmission in the system?
How is activity aggregated between single neurons and other neurons in the system?
How can one sufficiently model the interaction and function of biologically complex neu-
rons and neural assemblies?
2.1 CyToaRChITECTURE
Most of what is known about the role of neurons as the computational elements of the nervous
system has been derived by measuring the cause and effect of experimental situations on neuronal
firing patterns. From these initial functional observations, the neurons belonging to the neocortex
that generate movement programs were spatially divided into four lobes (frontal, temporal, parietal,
and occipital) and have later been associated with motor imagery [
1
], visual/tactile manipulation
1
Groups of neurons have a variety of means of organizing their inputs.
Convergence
refers to neural input from a
diverse set of sources, whereas
divergence
refers to a neuron's ability to exert influence on multiple target sites.
