Our general objective is to measure precisely how color and pattern information are jointly processed in human vision. Detection experiments for a test flash on a large field can isolate various chromatic pathways receiving inputs from different combinations of long (L), middle (M), and short (S) wavelength cones. For stimuli of greater than 520 nm only L and M need be considered. By exciting these in various ratios (positive and negative) we have measured threshold detection contours and have identified 2 separate mechanisms: luminance (Lum) responding to M+L stimulation and red-green chromatic (Chr) responding to the M-L difference. Using the paradigm of the pedestal (coincident in space and time with the test flash), the degree of suprathreshold interaction between these mechanisms is measured. Pedestals of varying strength and different mixture of Lum and Chr components are used. Preliminary results show that for 1 degree test stimuli on a yellow field, Chr detection is weakly facilitated by suprathreshold Lum pedestals and this facilitation persists to very intense pedestals. A kind of suprathreshold independence for Chr detection is thus suggested. Since physiological studies show that in early processing stages (e.g., ganglion cells), spectral tuning is strongly influenced by stimulus size, we will measure the interactions for small-scale stimuli including tests spots down to 5' and sine wave gratings of greater than or equal to 10 cycles/degree. The second thrust of our work is to incorporate S cone excitation. In the 3-dimensional space of S,M, and L stimuli, the threshold is a surface. A minimum of 3 linear mechanisms are required to describe it: presumably, (a) luminance, (b) red-green Chr, and (c) the putative blue-yellow (B-Y) opponent mechanism. If the principle of independence extends to 3-dimensions, the threshold surface will be a parallelepiped. We intend to test this conjecture with precise data. To perform this exacting task we use a computer-controlled 8-channel Maxwellian view system which can deliver positive or negative stimulus increments arbitrarily to the S, L, and M cones. Stimuli can also be given spatial and temporal patterns. By identifying discriminable psychophysical mechanisms, it is possible to identify analogous physiological substrates. These then become anatomical loci whose dysfunction can be discerned noninvasively.
Showing the most recent 10 out of 19 publications