 |
|
|
|
Early visual pathway
When light hits the eye, neurons in the retina on the back side of the eye
get excited. Several types of neurons are involved in the signal processing
in the retina, but the output signal which is transmitted towards cortex,
leaves from so called ganglion cells. The ganglion cell axons (which are long,
thin neuronal out-growths which propagate the signal to other neurons)
constitute the optical nerve which transmits visual information to a part of the brain called
lateral geniculate nucleus (LGN) which is a part of
thalamus. The main function of thalamus is to transmit sensory stimuli
from the outer sensory systems to cortex.
The so called relay cells in LGN receive visual signals from ganglion cells
and transmit processed information to the primary visual cortex.
From primary visual cortex the signals are then transmitted to other parts
of cortex, and this eventually results in a visual perception of the
surrounding world.
Receptive fields
An important notion in studies of the visual system is receptive field.
This term refers to the limited area in the visual field of the eye which
when stimulated by light (or darkness) influences the firing activity of
a neuron. The influence from an area inside the receptive field of a neuron can be both
excitatory and inhibitory. For ganglion cells the receptive fields are
small circular areas, and they exhibit so called center-surround
antagonism . So called ON-cells have their maximum firing activity
when they are stimulated by a circular spot of light (on a dark background) which
exactly covers the so called receptive-field center. [OFF-cells have opposite
response characteristics and responds to dark spots.] Illumination of an area
outside this receptive-field center, on the other hand, will reduce the firing activity.
Therefore, the receptive field can be described as a circular excitatory area surrounded by a
ring-shaped inhibitory area. This antagonistic center-surround organization makes the
system more suited to detect changes in the light intensity than the absolute magnitude of the intensity.
Mathematical modelling of receptive fields
The small and roughly circular receptive fields of cells in LGN have traditionally been modelled by
the difference-of-Gaussians (DOG) model. Here the receptive-field function (or point-spread function)
is modelled as a difference between two Gaussian functions, one representing a narrow excitatory center and the other
the wider inhibitory surround. The DOG-model is an example of a descriptive (or filter) model
where the purpose of is to summarize experimental data compactly in a mathematical form. In
mechanistic modeling one attempts to account for nervous system
activity on the basis of neuronal morphology, physiology and
circuitry. This corresponds to the
traditional physics approach to mathematical modeling of natural
systems.
Mechanistic modelling of receptive fields in LGN
The lack of detailed information about the neuronal circuitry has to a
large extent prohibited mechanistic modeling of neuronal circuits in
the visual pathway. During the last decades major progress has been
made in mapping out the properties of the neurons in the dLGN
and their synaptic connections. The LGN is a, relatively speaking, simple system
with few cell types and, compared to cortex, modest divergence and
convergence of synaptic connections. This limited complexity makes the
construction of mechanistic models less ambiguous and reduces the
number of unknown model parameters. Further, since a lot of physiological
experiments have been done on the LGN circuit, data are available for falsification of
tentative mechanistic models.
We have worked and are currently working on developing such mechanistic model for the receptive fields of LGN cells. In an early project we focused on the feedforward aspect of the LGN-circuit (i.e., feedforward excitation from retinal ganglion cells and indirect feedforward inhibition of LGN relay cells via intrageniculate interneurons). Later we have focused on the role of feedback on the response properties of LGN cells, with a particular emphasis on the cortical feedback. The work is done in close collaboration with the experimental group of Paul Heggelund at the Department of Physiology, University of Oslo.
Publications
Presentations