Using the distortion of visual probes to analyse human visual pathways

In human perceptual or 'psychophysical' experiments, we precisely define the visual stimuli that are presented to human subjects in terms of of temporal frequency, spatial frequency, wavelength and brightness. By manipulating the stimuli along some dimension, and measuring the change in the subject's response, we can infer something about the properties of the underlying visual system. Unfortunately, such measures of the intact/complete visual system do not grant us direct access to the processes that intervene between the stimulus and response. Physiologists, by contrast, have the advantage that they can make recordings within the (non-human) visual system by inserting electrodes directly into visual pathways. The use of visual nonlinearities goes some way to redress this imbalance. With them, the psychophysicist also has an 'electrode' with which to record and introduce signals within the visual pathways and so functionally 'dissect' the human visual system. In this project, we propose the novel use of visual nonlinearities to study the two primary human visual pathways, the chromatic and luminance pathways--from the inside. A visual nonlinearity is a stage in the visual pathway that distorts signals that pass through it to produce new signal components that were not present at its input. The two nonlinearities that we will study distort sinusoidal flicker signals to produce--in addition to flicker--a steady change either in colour or in brightness, which appear superimposed on the flicker and disappear when the flicker is turned off. For example, a red disc of light (c. 670 nm) can appear yellower when it is flickered (colour distortion) or a greenish-yellow disc (c. 565 nm) can appear brighter ('brightness enhancement'). Figures 1 and 3, below, illustrate these effects for flicker that slowly varies in amplitude ('amplitude-modulated flicker'). By carefully manipulating the stimulus presented to the subject, and thus the new signal components produced at the nonlinear site, it will be possible to measure separately the temporal frequency responses before and after each nonlinearity. Two nonlinearities will be investigated in the M- and L-cone pathways: one, which is compressive, causes flicker bursts to change in apparent colour and saturation, while the other, which is expansive, causes them to change in apparent brightness. Our expectation is that we will be able to determine separately the early and late properties of the two principal human postreceptoral pathways: the putative chromatic and luminance pathways, and thus provide new insights into the postreceptoral organization of the human visual system and the processing changes going on in them. The colour and brightness changes that we will investigate are distinct: first, because they can occur separately, and, second, because they are consistent with different types of nonlinearities (compressive or decelerating in the case of colour; expansive or accelerating in the case of brightness) and different percepts. Taken as a whole, these observations suggest the exciting possibility that the two nonlinearities are in the luminance and chromatic pathways, the two principal postreceptoral pathways in the human visual system. We believe that this project will provide a unique window into the internal workings of the human visual system, and will identify key signals and sites that will yield to physiological and/or anatomical study in other species.

Nonlinearities revealed by the visible distortion of amplitude-modulated (or other time-varying) flicker signals will be used to functionally dissect visual pathways in the human visual system into early (pre-nonlinearity) and late (post-nonlinearity) stages. Such nonlinearities provide a landmark at which to dissect the visual pathway, because they distort input signals internally to produce new signal components within the visual system. By carefully manipulating the visual stimuli presented to the subject, we can experimentally control the new signal components produced at a nonlinear site in ways that make it possible to determine separately the properties of the visual pathway before and after the nonlinearity. We will investigate two nonlinearities in the M- and L-cone pathways. Each causes bursts of flicker to be perceived differently from a steady light of the same time-averaged intensity and chromaticity. One, which is a compressive nonlinearity, causes bursts of M- or L-cone detected flicker to change in colour and saturation. The other, which is an expansive nonlinearity, causes bursts of M- or L-cone detected flicker to increase in apparent brightness. We will use each nonlinearity to dissect the visual pathway in which it resides. In some experiments, the nonlinearity will be used to measure the temporal frequency response of the pathway before the nonlinearity; while in others it will be used to measure the temporal frequency response of the pathway after the nonlinearity. Several methods of measurement will be used. We believe that the compressive nonlinearity lies in an M- and L-cone chromatic pathway, whereas the expansive nonlinearity lies in a separate M- and L-cone luminance pathway. Thus, we will be able to study separately the early and late stages of both the chromatic and the luminance pathways. The results of the proposed experiments will provide new insights into the postreceptoral organization of the visual system.