Temporal Characteristics of Monoptic, Dichoptic and Half-Binocular Collinear Lateral Masking of Contrast Detection

Purpose. Few prior studies have investigated the temporal properties of inter-ocular (i.e. dichoptic) contrast integration across space in primary visual cortex. My pilot study used collinear flanks to investigate the effect of varying the interstimulus interval (ISI) and flank duration on contrast detection threshold (CDT). As expected, the results revealed CDT facilitation at shorter stimulus onset asynchronies (SOA = ISI + flank duration) and reduced inter-ocular contrast integration relative to intra-ocular (i.e. monoptic) integration. It also showed unexpected, inter-ocular CDT suppression at longer SOAs. To better understand that surprising finding, I conducted additional experiments that added more subjects, longer SOAs, an additional viewing condition (half-binocular), and orthogonal flanks.;Methods. Eleven subjects with normal vision participated. Nine were naive to the purpose and participated for a mean of 25 hours each. The primary investigator (WK) and a summer research fellow (MM) participated for approximately 240 and 100 hours, respectively. In the main experiment, target and flanks were three cpd vertical sinusoids separated by six lambda (sigma=1.5 lambda) center-to-center vertical separation. Flank contrast was normalized to 3X flank CDT. Flanks were presented at four durations (67-500ms) and ISIs at seven durations (0-2500ms) resulting in SOAs from 0-3000ms. Target presentations were 250ms to the dominant eye via mirror haploscope and septum. Flanks were presented to dominant (monoptic and half-binocular viewing) and non-dominant eyes (dichoptic and half-binocular viewing). Forward masking was used with a 1-FC detection paradigm and 7-level MOCS. Each threshold was calculated from approximately 700 trials (approximately 10 runs over 2-3 days). A supplemental, orthogonal flank experiment resembled the main experiment with the exception of flank orientation and SOA range (0-1000ms).;Results. As expected, simultaneous presentation of collinear flanks resulted in mean CDT facilitation (monoptic 18.9% ± 3.9% (SE); dichoptic 13.9% ± 4.0%; half-binocular 18.0% ± 4.2%). For all viewing conditions, relative facilitation decreased as SOA increased up to 1000ms. Surprisingly, dichoptic and half-binocular viewing showed CDT suppression at long SOAs beginning at 500ms (dichoptic) and 750ms (half-binocular), with maximal suppression (9.9% ± 5.1% and 5.3% ± 4.7%, respectively) occurring at 1000ms. For dichoptic viewing, the CDT suppression was statistically significant (p < 0.05) at all 5 SOAs from 500-1000ms. All viewing conditions approached no effect at the longest SOAs (1500-3000ms). Flank duration had a significantly greater contribution to the overall effect than ISI for monoptic and half-binocular viewing. There was no significant difference in contribution under dichoptic viewing. Both monoptic orthogonal and dichoptic orthogonal flanks produced CDT facilitation at shorter SOAs that decreased with increasing SOA. Importantly, neither orthogonal flank condition produced CDT suppression.;Discussion. The collinear CDT facilitation produced by intra-ocular and inter-ocular flanks at shorter SOAs is consistent with the properties of long-range, lateral connections in primary visual cortex. This facilitation persists well beyond the maximal temporal integration limit (approximately 200ms) of the transducer model of contrast integration and therefore appears inconsistent with that model. The reduced degree of dichoptic CDT facilitation at shorter SOAs (compared with monoptic viewing) is evidence of decreased inter-ocular contrast integration. In general, the results are in agreement with existing models of intra- and inter-ocular contrast gain control. The temporal aspects of long SOA inter-ocular CDT suppression observed in the present study are consistent with the temporal properties of illusory contour perception reported in prior studies.;Conclusions. I propose the novel hypothesis that the CDT suppression produced by collinear flanks at longer SOAs under dichoptic and half-binocular viewing is due to one-way, contrast adaptation from lateral propagation that produced the effect of a collinear, illusory contour. This hypothesis is supported by the dichoptic, orthogonal flank experiment that showed no CDT suppression at the same longer SOAs.