The two main electrophysiological techniques that are used with human observers are the visually evoked cortical potentials (VECP) and the electroretinogram. The former consists of placing electrodes on the scalp and recording the gross electrical activity of the cortex beneath the electrodes. The electroretinogram involves placing a contact lens, with an embedded electrode, on the cornea. This permits the recording of the gross retinal electrical potential. There is yet one other electrophysiological recording method that can be used to evaluate the visual response. However, because it is a particularly invasive technique it is never used on human observers. This technique is called single unit recordings and involves inserting an electrode into a neuron (intra-cellular recording) or near a neuron (extra-cellular recording). Our major concern in this book is human vision, consequently I will focus more on the first two methods than the last.

Visually Evoked Cortical Potentials (VECP)
The electrical responses recorded with surface electrodes attached to the scalp over the occipital cortex and a reference electrode at some other convenient place some where on the head or ear.

The records obtained yield a rather complex wave form. Indeed, it is often very difficult to recognize any change in the wave form as a result of a visual stimulus. If one assumes that the shape of the recorded trace is random except for the location on the trace where the stimulus was presented, then if multiple cortical responses are averaged the location of the trace where the stimulus occurred should become obvious. This technique of averaging cortical responses to visual stimuli is done by an instrument called the computer of average transients (CAT). When this technique was first developed a specialized computer did the work. Now most computers can be programmed to handle the task. If I have confused you with this idea of averaging cortical Reponses, make sure you look at the graphic entitled averaging cortical responses.

Bieber et al. (1995) obtained VECP recordings from human infants and adults as they viewed heterochromatic flicker photometry stimuli. The humans also used the method of adjustment to psychophysically determine spectral sensitivity by means of heterochromatic flicker photometry.

To obtain a spectral sensitivity function from VECP data the first thing that is needed is the criterion VECP response at minimum flicker. The reciprocal of the radiance associated with the criterion VECP are then plotted as a function of wavelength to yield the spectral sensitivity function.

ElectroRetinoGram (ERG)
When light enters the eye it eventually is absorbed by photopigments in the outer segments of the receptors which initiates an electrophysiological sequence of events that make their way to occipital cortex of our brain. Above we discussed the recording of the electrophysiological activity from the cortex. Here we discuss the recording of the electrophysiological activity of the retina.

These recordings are made by embedding a corneal metallic electrode in a scleral contact lens and placing a reference electrode at some other convenient place (for example, the forehead). Such recordings are possible because of the good contact made with the retina by media surrounding the retinal tissue and the electrical contact these media have with the cornea.

A major component of the ERG comes from the activity of the rods. Since, at present, we are interested in the photopic spectral sensitivity, which depends on the cones, it is necessary to suppress this rod component. Aiba et al. (1967) used this approach to measure the photopic sensitivity function. They superimposed foveally fixated narrow band test stimuli on a bright blue background. This background saturated the rods and suppressed their contribution to the ERG. These investigators then measured the amplitude of the ERG as a function of the test stimulus intensity. These functions plotted on log-log coordinates yield straight lines. To obtain an action spectrum (spectral sensitivity function) they took a criterion ERG amplitude and determined the associated stimulus intensity the reciprocal of which could then be plotted as a function of wavelength. If the reader is confused, it would be a good idea to once again look at the ERG amplitude as a function of stimulus intensity function. The result is a photopic spectral sensitivity function based on ERG recordings. Aiba et al. compared these ERG based function with the subjects' psychophysically based spectral sensitivity function and with the CIE photopic spectral sensitivity function.

As can be seen the fit to the psychophysical function and to the CIE function is excellent for one observer and quite good for the other except at the short wavelength end of the spectrum. Unfortunately, these are the only data I've been able to find. There was one previous study in 1954 but while ok for its time, it is rather questionable by more current standards.

It has been shown rather conclusively that the a-wave of the ERG is a cone response. In principle it should be possible to determine a cone spectral sensitivity function based on the a-wave response. My limited search would seem to indicate that this has not yet been done. However, there is little reason to expect a result that differs from that obtained by Aiba et al. (1967).

Table of Contents
Subject Index
Table of Contents [When not using frames]