Contribution of area 19 to the foreground-background-interaction of the cat: an analysis based on single cell recordings and behavioural experiments

Summary

The contribution of area 19 to pattern discrimination in the cat was studied by single cell recordings in this area and by behavioural experiments before and after bilateral lesions. In order to make quantitative comparisons between behavioural performance and that of cell systems, we introduced a new parameter that characterizes visual neurons by their signal-to-noise (S/N) thresholds. A structured visual background made up of Gaussian visual broadband noise which could be moved was superimposed on the signal (moving bars or outline patterns) and the S/N characteristics of the response were determined by varying the signal intensity. The detection performance of cats after bilateral lesion of area 19 showed no deficits. Only for slowly (11 deg/s) or quickly (110 deg/s) moving patterns, or when the background was moved relative to stationary patterns, did we find slight, but significant deficits in the low S/N range. However, when the S/N ratios were higher than 5, all cats achieved their full preoperative performances and no deficits remained. The S/N thresholds of neurons in area 19 were much higher than those found for neurons in areas 17 and 18. The lowest thresholds were found with a stationary background. Introduction of relative velocity between background and bar resulted in intermediate thresholds and the highest thresholds were observed for stimulus configurations lacking relative velocity. These effects correspond to the performance of the intact animal, in which introduction of relative motion increases the performance. The S/N thresholds did not correlate with levels of spike rate recorded at high S/N ratios, direction selectivity or speed preference, indicating that S/N threshold measurements provide a significant additional description of visual neurons. A limited number of area 19 cells recorded in area 17/18 lesioned animals showed very similar thresholds suggesting that this property may be independent of the intactness of areas 17 and 18. The residual performance by 17/18 lesioned cats in detecting small patterns corresponds well to the characteristics of the single cells of area 19. This suggests that area 19 might be able to make a considerable contribution to this task when areas 17/18 are eliminated, though by itself it seems not to be able to sustain the level of performance mediated by them. The contribution of area 19 is restricted to performances at high S/N ratios only. In contrast to what was found for areas 17 and 18, area 19 makes no essential contribution to lowering the S/N ratio at which the system is able to detect the presence of a pattern in a background of irrelevant detail.