| |
Visual electrophysiologists have at present a wide choice of instruments at their disposal, from direct recorders to sophisticated computers. This enables them to record ever smaller responses and to improve their quality. “Thresholds,” defined as the weakest stimuli evoking recognizable responses, are continuously dropping, and the range and type of electroretinogram (ERG) and visual evoked potential (VEP) stimuli have been regularly extended. In addition, computer-based analytical methods are increasingly being used for the characterization of responses.
A proper selection from these modern methods requires knowledge about the principles of signal analysis on which they are based. These same principles apply to many quantitative aspects of visual function. The present chapter is meant to help the researcher and the clinician to find their way among the multitude of published methods. Emphasis will be laid less on mathematical rigor than on the understanding of fundamental concepts. The topics to be covered are (1) the recording and processing of electrical responses and (2) analytical questions concerning the stimulus and response characterization.
The unavoidable presence of noise (e.g., the background electroencephalogram [EEG]) demands procedures for noise reduction, especially if weak stimuli are presented. This can be done in a variety of ways, but in clinical practice it is often most important to reduce the recording time, which has encouraged the use of more “efficient” methods such as, for instance, the so-called steady-state stimulation. It is intended that the present chapter will enable the evaluation of such techniques.
For a long time nearly all ERGs and VEPs were recorded with flashes. The responses obtained in this way are often complex and more prone to the effects of strong nonlinearities (for a definition of this term, see below). However, by employing stimuli with other waveforms, among which sinusoidal modulation is the most frequent, it is easier to recognize deviations from linearity and to identify significant nonlinear properties of the system under study. An added advantage of sinusoidal stimuli is that linearity can often be approximated to a satisfactory degree, which facilitates analysis and description.
More recently homogeneous field stimulation has been superseded by the use of spatially structured fields such as checkerboards and sine wave gratings. This type of research has developed in two main directions:
1. Characteristics such as amplitude, wave shape, and latency are used to discriminate between normal and pathological responses. In this case only rather elementary methods of signal improvement need be employed.
2. The responses are used as a criterion for the “effectiveness” of a changing stimulus, e.g., when the size of checks in a checkerboard or the periodicity and the contrast of a grating are manipulated to obtain a constant response. The results of such studies are often expressed as a contrast sensitivity function (transfer function). The theoretical background, however, is complicated and requires among other things an analytical characterization of the stimulus, e.g., that of a checkerboard.
| |
