New BBSRC grant, August 2014

Multiple signals interact in flicker: recursive surround networks

Project summary:

An image of the outside world is formed by the lens and cornea onto the rear surface of the eye, the retina. The retina is carpeted by light-sensitive photoreceptors, called rods and cones, that convert the arriving photons into signals that that can be encoded and processed by neural circuitry in the retina and brain before finally reaching our conscious perception. There are three types of cones, which are responsible for daytime colour vision, called short-wavelength, middle-wavelength and long-wavelength, according to the part of the visible spectrum to which they are more sensitive (short-wavelength light usually appears violet and long-wavelength red). The main focus of this work is to understand more about how neural signals leaving the light-sensitive cones are processed and analysed by the human visual system before they are consciously perceived.

The signals from the cones are transformed and encoded by multiple circuits in the visual pathways through the retina and brain. These processes subtly alter how we perceive visual stimuli.  Many of the processes cannot be directly observed – they are carried out by small neurons with a large number of even smaller connectors. Together, these complex interconnections and interactions determine what we eventually perceive. One way of determining some of the characteristics of these inaccessible processes is to study how our behaviour depends on them. Our aim is to investigate the circuitry of the human visual system through measurements of how carefully designed changes in visual stimuli alter our perception of those stimuli. In this work we use mainly visual stimuli that vary in time (in intensity or chromaticity) and thus produce the perception of flicker.

We rely on the fact that interactions between different circuits within the visual system alter visual signals in characteristic ways. Typically, these interactions cause flicker to be relatively less visible at some frequencies but easier to see at other frequencies. By looking for those characteristic effects in perceptual measurements, we can tease apart the different underlying visual circuits and their contributions to perception. In previous work, we have discovered a surprisingly complex circuitry that uses the signals from the individual light receptors—cones—in several different and unexpected ways. For example, the neural circuitry can substantially delay the signals from one cone type relative to another, or change its sign, so that the two signals subtract rather than add. In this project, we propose to monitor and model the signals from different cones under a variety of new conditions to understand more about how the eye and brain work. We hope this work will provide a greater understanding of vision and visual processing, and will help to bridge the widening gap between the relatively simple models of the visual pathways derived from perceptual experiments and the complexity of structure and function of the visual system being revealed by other disciplines.

Example of a recursive surround network between cones with each step separated by delays (simple delays and exponential decays) and signal inversions.