Saturday, October 26, 2019
Study into Mapping Blind Spots
Study into Mapping Blind Spots Though the blind spot is a part of the retina that is devoid of photoreceptors, therefore relaying no visual information, it does not lead to the experience of a dark hole in our visual field (Sakaguchi, 2001). This is due to the perceptual phenomenon of filling-in, whereby a visual attribute such as colour and brightness is perceived in the blind-spot, even though it only exists in the surround (Komatsu, 2006). A long-standing question has been whether perceptual filling-in ignores this absence of information or actively fills it in (De Weerd, 2006). The symbolic theory proposes that early visual areas only extract information at the surface border with the colour and shape of the surface reconstructed in higher areas (Komatsu, 2006). Conversely, the isomorphic theory assumes that the retinotopic map in the primary visual cortex (V1) receives information from the entire surface with visual features such as colour activated in early visual areas (De Weerd, 2006). Hence Komatsu (2006) proposes an amalgamation of the theories, that neural activity is higher along the edge of the blind-spot with these signals then spreading across a two-dimensional array of visual feature sensitive cells in early visual areas. Hence the mechanisms of filling-in depend upon activity along the physiological edge of the blind-spot as Spillmann, Ottee, Hamburger and Magnussen (2006) found that a ring as thin as 0.5 Ã ¯Ã¢â¬Å¡Ã ° had been sufficient in inducing colour filling-in. Li et al., (2014) analysed this through 2.5Ã ¯Ã¢â¬Å¡Ã ° mono-coloured annuli, finding that it filled in completely, providing evidence for active colour filling-in from a small border. Fahle and Schmid (1987) proposed that the mean distance between photoreceptors is slightly higher for the temporal side compared to the nasal side in the retina with the ganglion cells distributed in a similar asymmetrical fashion. This spatial arrangement of the image as it is processed within the retina is maintained in V1 (De Weerd, 2006). Hence Li et al., (2014) extended the study of homogenous stimuli to bi-coloured rings, revealing the presence of a retinotopic rule in perceptual filling-in that favours the nasal side. Whilst this validates rapid colour filling-in as preattentive, these spatial arrangements may be affected by other preattentive factors which contribute to global processes such as relative salience (Brown Thurmond, 1993). Hence the current study aimed to examine ambiguities in perceptual filling-in when responding to both lower and higher-level processes. More specifically, whether differences in the relative salience of bi-coloured annuli affected the nasal or temporal processing (retinotopic rule) in filling-in blind-spot. As Brown and Thurmond (1993) infer that relative salience contributes to higher processes, exposure to increased saturation may alter the retinotopic rule. Based on these two studies, it was hypothesised that the average choice probability for asymmetrical filling-in would decline as the relative saturation for the nasal side decreases. Reciprocally, it was hypothesised than asymmetrical choice probability would increase as relative saturation for the temporal side decreases. Method Participants Fifteen students from the University of Sydney (4 men, 11 women; M age = 21, SD = 2.03), participated voluntarily. All participants had normal or corrected-to normal vision. With the exception of the three experimenters, participants were naÃ ¯ve to the experiment. Materials Stimuli: All stimuli was generated using Microsoft PowerPoint Software. Stimuli consisted of bi-coloured (red and green) annuli, with a width of 2.5Ã ° (derived from Li et al., 2014). For each participant, the diameter of the stimuli was adjusted so that the annuli overlapped with the edge of the blind spot. The two halves of the bi-coloured annuli were juxtaposed symmetrically on the nasal and temporal sides of the blind spot. Each side was counterbalanced across trials wherein half the trials were comprised of nasal-red: temporal-green stimuli and the other half, nasal-green: temporal-red. The saturation was adjusted for one side to 100%, 50% or 25% of the original saturation, while the other half was maintained at 100% saturation (that is, 100:100, 100:50, 100:25, 50:100, 25:100). Controls used reversed stimuli, such that the fixation cross appeared on the right and the stimuli on the left. Thus, 36 randomised trials were conducted, consisting of six controls and three repeats of ten test stimuli (Appendix A1). Choice Panel: This illustrated the spread of the two colours in coloured disks and consisted of ten choices (refer to Appendix A2). Procedure The experiment was conducted over two sessions, one week apart. In the first session, the blind spot of each participant was mapped using Microsoft PowerPoint. Participants were seated in a dark room with a chin rest at a distance of .57m away from an ASUS S400c 14-inch screen. Participants were instructed to fixate on a white fixation cross presented on a black background with their right eye and left eye covered. Using a digital pen tool, a small white test dot was moved across the screen by the experimenter. The positions where the dot disappeared and reappeared were verbally reported by the participant and digitally marked when it was not visible. The process was repeated until the blind-spot had been mapped out adequately. In the second session, participant were asked to report the perceptual filling-in of the blind-spot. The fixation slide (5 sec) and the stimulus slide was presented (3 sec). Participants were then presented with the choice panel and asked to report the choice that best resembled what they observed. At the completion of the study, participants verbally reported their experiences with filling-in (Appendix B). Results Paired sample t-tests were conducted, with participants reporting nasal colour dominance significantly more often than either symmetrical filling-in, t(1,14) = 2.37, p =.03 (nasal red: M = 40.1%) and t(1,14) = 3.09, p M = 51.2%), or temporal colour dominance t(1,14) = 5.79, p M = 60.1%) and t(1,14) = 9.13, p M = 75.6%). A 5 x 2, repeated measure analysis of variance was carried out on colour and relative saturation on average choice probability for nasal dominance filling-in, after which quadratic trend contrasts were conducted. Averaged over relative saturation, choice probability for nasal dominance filling-in was significantly greater when the nasal colour was green than red (mean difference = 21.4%), F(1,14) = 15.30, p F(4, 56) = 3.56, p = .01. Quadratic trend contrasts revealed that this effect increased significantly as the nasal relative saturation increased from 25% to 100%, but decreased as relative saturation decreased from 100 to 25 on the temporal side, F(1,14) = 19.92, p F(1,14) = 0.00, p = .99, nor was there a significant cubic trend, F(1,14) = .37, p = .55. Figure 1. Average choice probability for nasal dominance filling-in as function of type of relative saturation (N = 15). Discussion Participants reported asymmetrical nasal dominance filling-in significantly more than symmetrical or temporally dominant filling-in. This replicated preliminary findings by Li et al., (2014), that filling-in from the nasal side occupied a much larger region than filling-in from the temporal side. A colour effect was revealed with a preference for green over red whilst controlling for salience, which Hamburger, Prior, Sarris and Spillmann, (2005) account for through higher-level processing of colour information. They postulate that typical background colours in natural scenes, i.e. green and blue, fill-in more easily than colours that are attributed to the foreground i.e. red and yellow. Hence in future studies, more colour pairings including blue and yellow should be tested to validate this theory. Yet, this colour effect may also reveal that relative salience was poorly controlled for in this study, as Brown and Thurmond (1993) manipulated saturation, reporting that a more salient colour is favoured when filling-in. This is because the green provided more contrast than the red as Hamburger et al., (2005) suggests that perceptual qualities of surfaces, e.g., saturation, affect other properties such as contrast and luminance. This raises concerns in the current study in the me asure of relative salience, as is it unclear whether changes in saturation alone prompted a global process that overruled the local processes involved in filling-in. Hence, in future studies these visual characteristics need to be carefully controlled for to ensure that it does not have a confounding effect on salience. As predicted, asymmetrical choice probability declined as the relative saturation for the nasal side decreased. This concurs with Li et al., (2014) that the strength of colour filling-in is determined by the retinotopic rule, whereby the direction of filling-in is correlated to greater cortical projection on the nasal side. Furthermore, Fahle and Schmid (1987) contend that the nasal side has a lower contrast sensitivity threshold compared to the temporal side which implies that the receptors on the nasal side were more easily able to detect a change in saturation, which increased the relative salience of the temporal side. This offered access to greater filling-in, which decreased the nasal-preference for asymmetrical filling-in of the disk. This can be extended for future research by also examining the effect of an increase in relative saturation e.g., 150%, 200%, 300% which may have an additive effect by strengthening the retinotopic rule (Brown Thurmond, 1993). However, contrary to the hypothesis that the asymmetrical choice probability would increase as relative saturation for the temporal side decreased, nasal dominant filling-in decreased as saturation decreased on the temporal side. As mentioned, the perceptual qualities of surfaces interact as Komatsu (2006) implies that the brain needs to integrate lower level visual information such as colour and brightness and decode it at the retinotopic map. This signal is then transmitted to higher cortical areas to eventually generate surface perception. Cortical processing in these early visual systems are heavily biased toward the detection of local contrast in luminance, resulting from edges, which is necessary in surface perception (De Weerd, 2006). In the present study when saturation was decreased on the temporal side, it also changed the luminance of the green colour, making it more salient relative to the red, amplifying the edge between the two colours. Hence, considering this local pro cessing preference for variations in luminance, Sakaguchi (2001) contended that the physical edge that exists in the annulus can activate the neurons coding them, as the two colours differ in luminance. This increased the salience of the temporal side, permitting a greater percentage of the temporal side of the disk to fill-in. Another limitation of the present study is that a majority of participants verbally reported seeing a black spot mainly on the temporal side, implying that the annulus did not fill-in completely. Yet as Li et al., (2014) argues that colour perception processes are rapid and preattentive, this incomplete filling-in cannot be attributed to an inadequate fixation time. Rather, this can be explained by other methodological issues as Spillmann et al., (2006) attribute this partial filling-in to improper fixation and involuntary eye movements that displace the annulus relative to the blind-spot. Spillmann et al., (2006) highlight the significance of this partial-filling in effect in validating that filling-in is an active physiological process generated by a narrow edge at the blind-spot. However as this dark shadow was reported mainly on the temporal side it can be explained by the more sparse distribution of receptors on the retinotopic map resulting in weaker temporal processes. As the width of the annuli remained constant, the nasal side filled in better due to a denser distribution of receptors (Li et al., 2014). Hence future studies should consider the relative width of the stimulus to suggest a width for the temporal side of the annulus in order to achieve symmetrical filling-in. In summation, this study presents evidence for active neural processes in retinotopically organized lower order areas, but also a role for higher order cognitive factors such as surface description (De Weerd, 2006). In the future, studies should attempt to map the size of the activated brain area to endorse this retinotopic asymmetry during filling-in and the effects of relative salience on this symmetry (Li et al., 2014).