Detection performance of normal cats and those lacking areas 17 and 18: a behavioral approach to analyse pattern recognition deficits.

The ability of cats to discriminate between two geometrical outline patterns in the presence of superimposed Gaussian visual noise was tested before and after bilateral removal of cortical area 17 and parts of area 18. The detection probability PD was measured as a function of the signal-to-noise ratio for the parameters: noise bandwidth, spatial frequency content and rate of movement of patterns. In both normal and lesioned cats a broadband noise was found to be most effective in masking the large patterns while two other types of noise, a medium frequency noise and a high frequency noise had little or no masking effect. For recognition of the smaller patterns in normal cats the medium frequency noise was found to be more effective than the broadband noise. The performance of the lesional cats was disturbed severely at low signal-to-noise ratios and was significantly inferior to that of normal cats-especially for small patterns. However, at high S/N ratios and for large patterns the performance of the lesioned cats was comparable to that of normals while for the small patterns they reached PD values inferior to those of normal cats. It is concluded that although pattern recognition can be performed successfully by cats lacking areas 17 and 18, these cortical areas probably make an essential contribution to this function under natural conditions in two ways: because of the X-type input of area 17, they increase the acuity of the system by making it more sensitive to higher spatial frequencies, and they permit detection of patterns at much lower S/N ratios i.e. they lower the signal-to-noise ratio at which the system is able to detect the presence of a pattern in a background of statistical visual noise. The latter effect is not limited to the higher spatial frequencies but also affects the very low spatial frequencies which are normally used for pattern detection. Previous failures to demonstrate clear deficits in pattern discrimination after 17/18 lesions in cats may be attributed to the fact that the patterns presented for discrimination were not masked by visual noise. Movement of patterns led to a slight, but not significant improvement of the performance in both normal and lesioned cats, but the deficits found for stationary and moving patterns were more or less equal.