Menu

Tim Gollish

Neuroscience Symposium
Kamermans Group

The Neuroscience Symposia are organized weekly by the Netherlands Institute for Neuroscience. The presentations are given by researchers from the institute or by guest speakers. The title and content of the symposium is usually made known in the week prior to the presentation.

Colloquium room – Netherlands Institute for Neuroscience

4:00 pm – The versatile retina: From linear receptive fields to nonlinear feature extraction
4:45 pm – Discussion & drinks

 

Abstract

The neural network of the retina is the first stage of visual processing. It is a complex network of more than 50 cell types that segregates the visual input into about 20-30 parallel channels of information. These channels correspond to different types of retinal ganglion cells, which are the output neurons of the retina. Despite this complexity, the classical view of the retina’s function is that of a fairly simple filtering and transmission device, which provides some adaptation as well as contrast enhancement and otherwise passes on information about local illumination for processing by downstream brain areas. Over recent years, however, it has become apparent that many information channels in the retina do not comply with this simple picture. In this talk, I will discuss examples from the work in my lab that show how analyzing retina responses to naturalistic stimuli, such as natural images or eye-movement-like image shifts, can reveal complex, nonlinear processing in the retina. The nonlinear operations allow different types of retinal ganglion cells to represent specific visual features. A recently observed example is the sensitivity of certain ganglion cells to the recurrence of spatial stimulus patterns across eye movements. Furthermore, we find that populations of direction-selective retinal ganglion cells encode information about eye-movement-like motion signals in the activity differences within the population rather than independently in the activity of individual cells. These examples contribute to a new, emerging view of the retina as a versatile processing device that – besides linear filtering, adaptation, and contrast enhancement – also contains nonlinear functions, which allow the extraction of specific visual features by populations of retinal ganglion cells.