A NOVEL AND ROBUST METHOD FOR ANALYZING VISUAL THRESHOLD DATA THAT APPEARS TO PROBE THE PROPERTIES OF PSYCHOPHYSICAL CHANNELS

Title A NOVEL AND ROBUST METHOD FOR ANALYZING VISUAL THRESHOLD DATA THAT APPEARS TO PROBE THE PROPERTIES OF PSYCHOPHYSICAL CHANNELS
Author, Co-Author Vincent Billock, Thomas Harding
Topic
Year
1992
Day
Sunday
Program Number
2:15 pm
Room
Ireland B
Affiliation
Abstract Billock and Harding (ARVO, 1992) reported that the correlation coefficient for contrast sensitivity thresholds is a monotonically decreasing function of the frequency separation of the pairs being correlated. This would be predicted by multiple channel models of spatial and temporal frequency processing; widely separated frequencies are detected by independent channels and should have less correlated thresholds than closer frequencies. We measured contrast sensitivity thresholds in 40 adults at 104 combinations of spatial and temporal frequencies. CRT and surround illuminance was 100 lumens; viewing was binocular with natural pupils. For temporal frequencies, we found that the correlation between thresholds was a linear function of frequency separation (7 octaves). For spatial frequencies, the correlation was linear for about 4 octaves, then leveled off. The slope of these functions is used as an index of processing -- high slopes indicate many channels with narrow bandwidths, low slopes indicate few channels with large bandwidths. The method is remarkably robust. When plotted with our method, data from other labs with similar spatiotemporal content have very similar slopes, despite differences in subject number, subject age, luminance, and psychophysical method. The data indicate: (1) more spatial channels than temporal channels; (2) more temporal channels at low spatial frequencies than high; (3) more spatial channels at low temporal frequencies than high; (4) evidence for more than two temporal channels. The method may be useful for exploiting the vast amount of normative data taken in clinical labs and for comparing data taken under different circumstances (or different disease states) for similarity of cortical processing.
Affiliation of Co-Authors
Outline