Experimental work on kittens has demonstrated that a normal visual environment is required to develop and maintain the special response characteristics of cells in the visual cortex. It is therefore very tempting to extrapolate these findings and explain the defects in human amblyopes by the same abnormalities in the visual environment. However, direct analysis of the function of the visual cortex in patients relies on indirect methods of investigation: because of practical difficulties, refinements of psychophysical testing (Campbell, Kulikowski, and Levinson, I966; Kulikowski and King-Smith, 1973) have rarely been employed in clinical work. Although, in the analysis of cortical defects, such investigations may highlight the abnormality and enable a precise description of the defect to be given (e.g. Bodis-Wollmer, 1972) in psychophysical terms, extrapolation from the psychophysical description to the cellular mechanisms remains very difficult. Another line of approach is to record cortical activity by using evoked potentials. Some studies of cases of amblyopia have been made with such a technique (Spekreijse, Khoe, and van der Tweel, 1972) and in these, as in normal subjects, correlations have been established between evoked potentials and pyschophysical sensitivity. Such analogies exist between various perceptual functions and their electrical analogues, e.g. binocular vision, suppression of visual acuity and colour vision (see Regan, I972, for a summary), but in such investigations the stiniulus has to be very carefully controlled and its production requires elaborate apparatus while the analogy can be drawn only after prolonged experimentation. For some years we have been using visually evoked responses (VERs) as an objective test of eye function (Behrman, Nissim, and Arden, 1972), using as a stimulus a patternreversing chequerboard. In order to make our test of practical clinical use, the testing procedure and analysis have had to be simplified, and we have used much the same testing procedures as before in the study of various types of amblyopia. Some preliminary results are described in this paper. In assessing their significance and their relationship to singlecell neurophysiological experiments, the limitations of the techniques must be borne in mind. The voltages recorded are obtained from relatively large areas of cortex and there is in general no assurance that in abnormal persons the responding cortical area is identical to that in the normal subject: alterations in the test parameters may also change the cortical area responding. One important illustration of this is the fact that changing the retinal location of the stimulus may change the polarity and the waveform of the recorded response (Halliday and Michael, 1970; Jeffreys, 197I). Since, in our technique, we measure both response amplitude and stimulus-response time, this fact may strongly affect the results. Another important variable is that the stimulus, being designed to maximize the response, will therefore stimulate various classes of co...

No abstract

The VER was studied in normal children aged 4--11 years and in 71 amblyopic children at Moorfields Eye Hospital. Changes in the VER in amblyopia are best observed with 20% contrast and small square sizes. The better eye of untreated amblyopia has a normal time-to-peak for P1 of the cortical response, while that of the amblyopic eye is delayed, the more so with poor binocular vision in partial agreement with our previous findings. We found with a high degree of significance that occlusion therapy renders the VER from the amblyopic eye more 'normal' but causes considerable delay in the VER from the covered eye, even though its visual acuity does not change. At the end of occlusion, these changes are usually reversible although not always completely in patients who have had prolonged occlusion.