Free
Research Article  |   February 2010
The effect of normal aging on closed contour shape discrimination
Author Affiliations
  • Allison M. McKendrick
    Department of Optometry & Vision Sciences, The University of Melbourne, Parkville, Australiaallisonm@unimelb.edu.au
  • Anne E. Weymouth
    Department of Optometry & Vision Sciences, The University of Melbourne, Parkville, Australiaanneew@unimelb.edu.au
  • Josephine Battista
    Department of Optometry & Vision Sciences, The University of Melbourne, Parkville, Australiajbat@unimelb.edu.au
Journal of Vision February 2010, Vol.10, 1. doi:10.1167/10.2.1
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      Allison M. McKendrick, Anne E. Weymouth, Josephine Battista; The effect of normal aging on closed contour shape discrimination. Journal of Vision 2010;10(2):1. doi: 10.1167/10.2.1.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Our experiments explore whether contour processing of closed shapes is altered by healthy aging. Contour processing was measured using a closed contour (circle or ellipse) constructed of Gabor elements. The contour was presented either on a blank background or embedded in noise (identical Gabor elements of random orientation). Twenty-one older (age range: 61–80 years) and 21 younger (age range: 22–38 years) adults participated in three experiments: 1) the number of Gabors comprising the contour was fixed (10, 12 or 15) and the threshold aspect ratio required to discriminate the shape (circle versus ellipse) was measured; 2) orientation jitter was added to the Gabor elements comprising the contour and shape aspect ratio discrimination thresholds were measured; and 3) the aspect ratio was fixed (three times the individual threshold aspect ratios) and the threshold number of elements required to determine the shape was measured. Older adults required a larger number of elements to discriminate the global contour shape ( F(1, 41) = 15, p < 0.001), even when stimulus saliency was matched for contrast sensitivity and aspect ratio threshold. This finding is consistent with other recent work showing deteriorations in cortically mediated visual processing with age.

Introduction
Many aspects of visual processing are altered by healthy normal aging (for review see: Sekuler & Sekuler, 2000; Spear, 1993). There are well studied changes to the optics of the eye (Savage, Haegerstrom-Portnoy, Adams, & Hewlett, 1993, Weale, 1988, Winn, Whitaker, Elliott, & Phillips, 1994), as well as substantial evidence for alterations in visual processing that can not be explained by optical deterioration alone. Examples include: alterations of motion perception (Betts, Taylor, Sekuler, & Bennett, 2005), center-surround contrast perception (Karas & McKendrick, 2009), bilateral symmetry detection (Herbert, Overbury, Singh, & Faubert, 2002), and the integration of local orientation information across space (Del Viva & Agostini, 2007; Roudaia, Bennett, & Sekuler, 2008). 
An important task performed by the human visual system is the detection and identification of objects. In order to meaningfully interpret visual objects, it is necessary for the visual system to integrate separate features into global shapes, and to segregate objects from their backgrounds (Loffler, 2008). This multi-stage process initially requires accurate encoding of local orientation information. A neurophysiological study of aged primates has demonstrated a reduction of orientation selectivity of neurons in primary visual cortex (Leventhal, Wang, Pu, Zhou, & Ma, 2003). A similar process in healthy human aging might result in deterioration of the ability to encode local orientation information. Human behavioral data does not support a significant decrease in orientation selectivity in older adults (Betts, Sekuler, & Bennett, 2007; Delahunt, Hardy, & Werner, 2008; Govenlock, Taylor, Sekuler, & Bennett, 2009). 
Intermediate stages of shape perception require the visual system to integrate separate local orientation features into contours (for review see: Loffler, 2008). A key research method for exploring the human visual system's ability to integrate local features into shapes is the contour integration task initially described by Field, Hayes, and Hess (1993). Contour integration tasks typically study the ability to detect contours comprised of local elements embedded in cluttered backgrounds. The task usually involves the detection of a line (snake) or shape that is comprised of local features such as line elements or Gabor patches. Precise alignment of elements along the local path is critical to task performance, as is the distance between the elements. These critical requirements are considered to reflect processing by long-range horizontal connections formed by pyramidal cells in V1 that link cells that have similar orientation preferences but non-overlapping receptive fields. The involvement of V1 in contour integration is supported by physiological evidence for facilitation of V1 responses by collinearly arranged line segments (Li & Gilbert, 2002; Li, Piech, & Gilbert, 2006; Polat, Mizobe, Pettet, Kasamatsu, & Norcia, 1998); although top-down influences are also important to learning this task (Li, Piech, & Gilbert, 2008). We use a variant of contour integration in this study. 
Two other studies have studied aspects of contour integration in the elderly. An initial study was performed by Del Viva and Agostini (2007) who used closed contours as stimuli. They measured the threshold number of noise elements required to correctly identify the position of a closed circular contour presented in noise and found that fewer noise elements were required to impair sensitivity for older than younger adults. Del Viva and Agostini (2007) did not find a relationship between contrast sensitivity and contour integration performance. The magnitude of impairment due to the addition of noise elements was similar across all contour inter-element distances tested. Stimuli were presented for 1 second potentially allowing some searching within the stimuli. Given that the results demonstrated a consistent decrease in performance across all stimulus conditions tested, performance potentially might be explained by an age-related difference in search ability or other high-level cognitive factor. Another possible confound with the Del Viva study is that the contours were constructed of elements of constant spatial phase. Contours constructed in this fashion can be detected by simple linear filters hence don't necessarily involve contour integration processes (Hess & Field, 1999). 
A second study of elderly contour integration was performed by Roudaia et al. (2008) who demonstrated that older people required higher contrast C shaped contours [comprised of local elements that were presented on a blank background] in order to correctly discriminate the position of the gap within the C. Younger observers showed an interaction between contrast and local element orientation such that contours constructed of tangentially aligned elements could be detected at lower contrasts than those comprised of radial (orthogonal to the C contour) or of mixed orientation. This same relationship between contrast threshold and element orientation was not present for the older observers. The authors interpret the lack of apparent facilitation in contrast threshold for the tangentially aligned stimulus as an age-related decline in contour integration. The study of Roudaia et al. (2008) doesn't investigate whether segregation of contours from noise is altered with aging, and is a little difficult to interpret given the known alterations in both threshold and suprathreshold contrast processing with aging (Karas & McKendrick, 2009; Owsley, Sekuler, & Siemsen, 1983). 
This study explores how aging alters the ability to discriminate between closed contours (circles and ellipses) comprised of local elements. We utilize methodology similar to that of Levi, Yu, Kuai, and Rislove (2007) who explored deficiencies in closed contour integration in amblyopia. In contrast to the study by Del Viva and Agostini (2007), we alternate the phase of the elements (to avoid cues that can be locally linked), have a shorter viewing time (to avoid eye movements and searching for the contours), and measure performance both in the presence of background clutter and without. The judgment required by observers was shape discrimination: ellipse or circle. The stimuli enabled us to assess contour shape discrimination while being able to manipulate local features of the contour—to study the effects of such manipulations and to balance saliency between groups. 
Our experiments were designed to answer the following questions: 
  1.  
    do older adults show a deficit in the ability to discriminate the shape of contours constructed of local elements?
  2.  
    if present, can the deficit in discriminating closed shape contours be explained by difficulty in encoding the local orientation of the elements?
  3.  
    does normal aging alter the inter-element distance required to correctly integrate local elements into closed shapes?
Methods
Participants
Our methods were similar to those described by Levi et al. (2007) who explored the effects of amblyopia on closed contour discrimination. Hence, we used the control group performance from that study to determine an a-priori sample size of 20 participants in each group [assuming a power of 0.8 to detect a difference in mean performance of older and younger groups on the shape discrimination task—data from their Figure 3]. Participants were recruited via written advertisements throughout the University of Melbourne and local community newspapers. Participant eligibility depended on normal findings from an eye examination designed to rule out ocular disease or significant media opacity. Inclusion criteria included visual acuities better than 20/30, intraocular pressures below 22 mmHg, as well as normal slit lamp, macula, and optic nerve head appearance. Refractive errors were required to be <±5 D sphere and <±2 D astigmatism. Our final sample included 21 older (15 female and 6 male; mean age 68; range 61–80) and 22 younger (15 female and 7 male; mean age 27; range 22–38) participants. 
All experiments were approved by our institutional Human Research Ethics Committee and complied with the tenets of the Declaration of Helsinki. All participants provided written informed consent prior to participation. 
Stimuli and procedures
Stimuli were presented on a gamma-corrected 21-inch monitor (resolution: 1408 × 1056 pixels; frame rate: 70 Hz; maximum luminance: 95 cd/m 2; G520 Trinitron; Sony, Tokyo, Japan). Custom software was written in Matlab 7 (Mathworks, Natick, MA, USA) interfaced with a ViSaGe system (Cambridge Research Systems, Ltd., Kent, UK). 
The monitor was viewed binocularly from 1 m using a chin and forehead rest. Participants were optimally refracted for viewing distance and wore corrective lenses in a trial frame where necessary. Participant responses were communicated via a button box (model CB6, Cambridge Research Systems). 
Stimuli were closed contours of Gabor patches (Gaussian windowed sinusoidal gratings) that were aligned to form circles (radius of 4 degrees of visual angle) or ellipses (see Figure 1). Individual Gabor elements had a spatial frequency of 1.5 c/deg and an envelope standard deviation of 0.33 deg. All contours had approximately the same geometrical areas, were formed by a variable number ( N) of elements, and were centered on the screen. The contour elements were placed with a random starting point and the starting spatial phase of the elements was alternated to be 0 then 180 deg. Contours were either presented on a gray achromatic background (45 cd/m 2) (see Figure 1A) or in noise. The noise comprised a full field of Gabor patches of random orientation but the same size and spatial frequency as those comprising the contour (see Figure 1C). To achieve approximately even density across the screen, the full screen area was divided into a grid of 18 by 14 (each grid approximately 1.25 degrees of visual angle). A Gabor patch was placed in each grid with the center of the Gabor being randomly positioned within ±0.5 grid size (±0.625 degrees) from the grid center both horizontally and vertically. Contour elements replaced noise elements in the same grid square. Noise elements were randomly allocated a spatial phase of either 0 or 180 degrees and noise grids were randomly computed for each presentation. 
Figure 1
 
Examples of closed contour stimuli: A) a circular contour constructed of 10 elements; B) an elliptical contour constructed of 10 elements; C) a circular contour of 15 elements embedded in a noise field.
Figure 1
 
Examples of closed contour stimuli: A) a circular contour constructed of 10 elements; B) an elliptical contour constructed of 10 elements; C) a circular contour of 15 elements embedded in a noise field.
For all experiments, participants were required to discriminate between a circular and an elliptical contour in a temporal two-interval forced choice stimulus presentation paradigm. Two presentations, one containing a circle (aspect ratio = 1) and the other containing an ellipse (aspect ratio > 1) were displayed for 200 ms each in random sequence, separated by a 500 ms interstimulus interval during which a blank screen of mean luminance was presented. Correct and incorrect responses were indicated with auditory feedback. The long axis of the ellipse was randomly oriented on each trial to prevent predictable positional cues being available to observers. 
Observers participated in three experiments. Experiment 1 fixed N (the number of elements in the contour: 10, 12 or 15) and varied the ellipse aspect ratio to determine a threshold aspect ratio for contour discrimination. Thresholds were estimated using a ‘three down, one up’ staircase (converging to the approximate 79% level: (Wetherill & Levitt, 1965) where the aspect ratio was varied in 20% increments at each staircase reversal. Staircases terminated after 4 reversals with the average of the last two reversals being taken as the staircase result. The final threshold estimate for each test condition was taken as the average of the results of two randomly interleaved staircases. Experiment 2 similarly measured the threshold aspect ratio for shape discrimination, but added random orientation jitter to the alignment of the elements along the contour path. In Experiment 3, we fixed the aspect ratio (to three times each observer's individual threshold) and varied the number of elements (N) to measure the threshold number of elements required to discriminate the shape. The same staircase procedure was implemented however the staircase step-size was one element at each reversal. Each experiment was conducted with and without the presence of noise elements. 
It is well established that aging affects contrast sensitivity (Owsley et al., 1983). To minimize any effects of differences in stimulus visibility between groups, before conducting the main experiments, we first measured contrast thresholds for the global contour stimulus. Contrast thresholds were determined for a circular contour (10 Gabor elements) displayed for 200 ms on the mean luminance background. A yes-no procedure was used whereby participants responded by pressing a button each time they saw the stimulus. The staircase procedure was the same as for the other experiments with the staircase step-size being a 20% variation in contrast. Two interleaved staircases were repeated resulting in the final estimate being determined from the mean of four staircases. For all subsequent experiments, stimuli were presented at five times the participant's contrast threshold. A contrast multiplier of five was chosen to ensure that the Gabor elements were clearly salient for all observers, but to avoid ceiling effects. The highest contrast threshold in the older group was 12%. 
Data was analyzed using SPSS 17.0 (SPSS Inc, Chicago, USA). A p-value of less than 0.05 was considered statistically significant. P-values are reported numerically throughout except for very small probabilities where a criterion judgement of p < 0.001 was applied. 
Results
Contrast threshold
As expected, the mean contrast threshold was elevated for older participants ( Figure 2, mean ± SEM younger vs. older: 5.4 ± 0.3 vs. 7.8 ± 0.5, Mann–Whitney U = 52, p < 0.001). 
Figure 2
 
Mean (±95% confidence intervals of the mean) contrast thresholds for younger (blue) and older (red) participants.
Figure 2
 
Mean (±95% confidence intervals of the mean) contrast thresholds for younger (blue) and older (red) participants.
Experiment 1: Ellipse aspect ratio threshold
This experiment measured the threshold aspect ratio for discriminating a circular contour from an elliptical contour for contours of fixed element number (10, 12 or 15). This experiment served two purposes: 1) to determine whether older adults have elevated closed contour shape discrimination thresholds, and 2) to determine whether the presence of noise has a greater impact on closed contour shape discrimination performance for older observers. The sequence of testing (number of contour Gabors) was randomized between participants, and all observers completed the no-noise condition first. 
Group performance is summarized in Figure 3 which shows ellipse aspect ratio thresholds for the no-noise (open symbols) and noise (closed symbols) conditions. There was a trend for the threshold aspect ratios of the older participants to be elevated compared to the younger participants ( Figure 3, two-way RM ANOVA, F(1, 41) = 3.5, p = 0.07). Shape discrimination was easier for all participants in the absence of noise ( F(1, 41) = 227, p < 0.001) and with more contour elements ( F(2, 82) = 50, p < 0.001). There was also an interaction between noise condition and the number of contour elements ( F(2, 82) = 52, p < 0.001) with thresholds increasing rapidly in noise. The difference between groups was independent of noise condition (no significant interaction between group and noise: F(1,41) = 2.5; p = 0.12). 
Figure 3
 
Mean (±95% confidence intervals of the mean) ellipse aspect ratio thresholds for varying numbers of elements comprising the contour for younger (blue) and older (red) participants. Results for stimuli presented within background noise are represented by the closed symbols (Panel A). Open symbols correspond to the condition without background noise (Panel B).
Figure 3
 
Mean (±95% confidence intervals of the mean) ellipse aspect ratio thresholds for varying numbers of elements comprising the contour for younger (blue) and older (red) participants. Results for stimuli presented within background noise are represented by the closed symbols (Panel A). Open symbols correspond to the condition without background noise (Panel B).
Experiment 2: Aspect ratio thresholds with contour element orientation jitter
Experiment 1 showed a trend for a small shape discrimination deficit in the older group. We were interested in whether this trend results from older observers having elevated internal noise in their encoding of the orientation of the local feature elements. We used a method previously implemented to explore the role of internal orientation noise in contour detection deficits in amblyopia (Hess, McIlhagga, & Field, 1997). Orientation jitter was added to the local elements and a subset of Experiment 1 repeated. We expected observer performance to decline when the stimulus orientation jitter was greater than their internal noise level for orientation coding. Hence, if older observers have elevated internal orientation noise, their thresholds should be less affected by small amounts of orientation jitter than younger observers. 
For Experiment 2, we fixed the number of Gabor elements to 15. The orientation of the local elements was selected at random from a Gaussian distribution of mean equal to the veridical contour orientation. Jitter was increased by enlarging the Gaussian standard deviation. Aspect ratio thresholds were measured in separate runs for Gaussian orientation jitter of standard deviations 5, 10, 15 and 20 degrees. 
Results
As in Experiment 1, mean ellipse aspect ratios showed a trend to be consistently elevated in the older participants, both with and without background noise ( Figure 4, two-way RM ANOVA, F(1, 41) = 3.8, p = 0.06). Shape discrimination was more difficult for all participants in background clutter ( F(1, 41) = 192, p < 0.001) and with orientation jitter ( F(4, 164) = 40, p < 0.001). The presence of orientation jitter was greater when the stimuli were presented within background noise (noise condition × orientation jitter F(4, 164) = 32, p < 0.001). The interaction between orientation jitter and group was significant ( F(4, 164) = 2.6, p = 0.04). Inspection of Figure 4 shows that the significant interaction is driven by the similar performance between groups for the highest jitter level tested. The data do not support the prediction that younger observers would be more affected by low levels of orientation jitter than older observers. 
Figure 4
 
Mean (±95% confidence intervals of the mean) ellipse aspect ratio thresholds for varying amounts of Gabor patch orientation jitter for younger (blue) and older (red) participants. Panel (A), closed symbols, corresponds to the condition with background noise and panel (B) shows the results for stimuli presented on a blank background (open symbols). The data for sigma = 0 degrees are replotted from Figure 3, 15 elements.
Figure 4
 
Mean (±95% confidence intervals of the mean) ellipse aspect ratio thresholds for varying amounts of Gabor patch orientation jitter for younger (blue) and older (red) participants. Panel (A), closed symbols, corresponds to the condition with background noise and panel (B) shows the results for stimuli presented on a blank background (open symbols). The data for sigma = 0 degrees are replotted from Figure 3, 15 elements.
Experiment 3: Number of elements required to discriminate closed contour shape
This experiment was designed to explore whether the distance over which local elements can be linked into contours differed between older and younger groups. We measured the minimum number (N) of local contour elements necessary for contour shape discrimination. Elements were placed at even distances around the contour, hence fewer elements resulted in greater inter-element distance. The task was the same as for previous experiments (a two-interval forced choice paradigm where the observer chooses the interval with the ellipse), however the staircase varied N. In order to balance ellipse saliency between groups, for each individual the stimuli were presented at 5 times contrast threshold (as in Experiments 1 and 2) and the aspect ratio of the ellipse was set a factor of 3 greater than the threshold aspect ratio determined for 15 Gabor elements in Experiment 1. Three times the threshold aspect ratio was clearly elliptical (pilot data demonstrated this to be on the upper asymptote of the psychometric function of normal observers, data not shown) but not sufficiently distorted from circular to result in some local elements being moved to markedly more eccentric retinal locations. 
The threshold number of elements required to discriminate the closed contour shape is shown in Figure 5. Older participants required more elements than younger participants, for both noise conditions ( F(1, 41) = 15, p < 0.001). Background clutter was more detrimental to contour discrimination for the older participants (noise condition × group F(1, 41) = 5.2, p = 0.03). It should be noted that when there is no noise present, with minimal elements (4), the task is no longer a shape discrimination task, but an inter-element distance judgement. 
Figure 5
 
Mean (±95% confidence interval of the mean) threshold number of contour elements for shape discrimination for younger (blue) and older (red) participants, with (closed symbols) or without (open symbols) background noise. The right y-axis refers to the noise condition. The axes have been split to enable visualization of the magnitude of the significant interaction between noise condition and group.
Figure 5
 
Mean (±95% confidence interval of the mean) threshold number of contour elements for shape discrimination for younger (blue) and older (red) participants, with (closed symbols) or without (open symbols) background noise. The right y-axis refers to the noise condition. The axes have been split to enable visualization of the magnitude of the significant interaction between noise condition and group.
Discussion
Our key finding is that older adults required more elements than younger adults to discriminate closed contour shapes ( Figure 5). The deficits cannot be explained by differences in contrast sensitivity since we approximately equated the contrast saliency of the stimuli between all individuals. We also balanced the stimuli for aspect ratio threshold differences between individuals. 
Our experiments studied contour integration for closed rather than open contours. The circle versus ellipse configuration has been used by others (Kuai & Yu, 2006; Levi et al., 2007), primarily to minimize the use of local orientation cues for contour detection (at threshold, the local curvature differences around the perimeter of the circle and ellipse are minimal). Closed contours had several advantages for our purposes. Firstly, contour integration performance is facilitated for closed versus open contours (Kovacs & Julesz, 1993; Mathes & Fahle, 2007). Primarily, closure appears to facilitate detection of the stimuli, such that closed stimuli are more salient than open contours. This ‘good Gestalt’ property is potentially advantageous when testing relatively untrained observers. Pilot data with our own stimuli demonstrated an approximate 30% improvement in aspect ratio threshold for the 15 element closed contour (data not shown). Removing one element keeps the local cues essentially identical to that of the 15-element condition, in particular the local curvature cues created by orientation differences around the radius of the circle or ellipse are intact. This pilot experiment was not intended as a robust investigation of closure versus non-closure as such experiments have been thoroughly conducted elsewhere (Kovacs & Julesz, 1993; Mathes & Fahle, 2007), however it is consistent with previous reports. The closed stimulus arrangement was also readily relatable to previous studies of shape discrimination (Wang, 2001) and circular contour integration (Del Viva & Agostini, 2007; Roudaia et al., 2008) in the elderly. 
In two experiments, our data demonstrate a trend for older adults to have a small deficit in the ability to discriminate closed contour shapes ( Figures 3 and 4). The group difference approached statistical significance ( p = 0.07 in Experiment 1, and p = 0.06 in Experiment 2). Prior to commencing the project we performed an a-priori power analysis that led to a planned sample size of 20 in each group. We tested an additional three observers (2 younger, 1 older) that were readily available. The data presented in Figures 3 and 4 suggest a small difference between groups, with an average effect-size (Cohen's d) of approximately 0.40. Assuming that the differences between the means and standard deviations presented in Figure 3 are representative of a larger sample, a sample size calculation estimates that 80 observers in each group would be required for a statistical power of 0.80. We interpret our data as strongly suggestive of a difference in aspect ratios thresholds between groups, however, the magnitude of the difference is small. 
There are few reports on the effects of aging on shape discrimination. Wang (2001) measured the ability of older observers to detect deformations from circularity for radial frequency (RF) patterns. Participants were permitted to view the stimuli as long as required before making their shape discrimination judgement within a spatial two-alternative forced choice paradigm. The older observers spent slightly longer making their judgements and demonstrated a small reduction in the ability to discriminate RF patterns for higher radial frequencies. Wang (2001) concludes that RF shape discrimination is largely intact with aging. Our data are reasonably consistent with this viewpoint, displaying a possible trend for a small elevation of aspect ratio thresholds when stimuli are presented briefly. 
Neurophysiology of aged primates demonstrates a decrease in orientation selectivity of neurons in V1 (Leventhal et al., 2003), however, human psychophysical data does not support a behavioral consequence of such changes if present within the human system (Betts et al., 2007; Delahunt et al., 2008; Govenlock et al., 2009). Our experiments similarly do not support a significant difference in the ability of younger and older observers to encode the local orientation of the elements comprising the closed contours. Less accurate local orientation encoding by older observers predicts orientation jitter to detract from performance more for the younger than older observers. This prediction was not supported by the data of Experiment 2. Instead, we show a similar fall off in performance of the two groups, except for the highest level of jitter (sampled from a Gaussian of mean 20°) when the groups perform equivalently. This leveling of performance at the highest level of jitter might partially reflect a ceiling effect whereby both the circular and elliptical contour were difficult to segregate from the background prior to actual shape comparison. Our data are consistent with that of Kuai and Yu (2006) who show that contour integration of closed contours is robust to jitter of about 12 degrees, which they suggest to be a correlate of orientation tuning of the neighboring spatial filters involved in contour integration. If our older observers have a broadening of such filters, we would expect our older observers to be more tolerant to intermediate levels of jitter (5, 10, and 15) than younger adults. This is clearly not the case in Figure 4
Experiment 3 demonstrated that a significant increase in the number of elements (or decrease in interelement spacing) was required by the older observers to discriminate between a circular and elliptical closed contour, both with and without background clutter. Our findings do not suggest a generalized problem with distinguishing contours from their background in our older observers because the detrimental effect of adding background noise was the same for older and younger observers in Experiments 1 and 2. Rather, our data suggest that older observers cannot perform inter-element linkages over the same distances as younger observers. 
An alternative possibility is that older observer performance reflects some form of cognitive aging factor, rather than purely perceptual factors. Of particular importance might be less efficient acquisition of information, differences in attention to the task, or differences in the ability to learn to perform the tasks. Our experimental design can not rule out these factors. We deliberately kept the stimulus presentation brief (200 ms) to minimize the capacity of all observers to search for the contours. Our older observers were all healthy, active and independent and were trained on the tasks, until the examiner was satisfied that the tasks were being performed correctly. Rest breaks were taken regularly for all participants. 
The neural mechanisms underpinning contour integration are not entirely understood. A number of studies have identified long-range horizontal connections formed by pyramidal cells in V1 that link cells of similar orientation preferences but non-overlapping receptive fields (Mitchison & Crick, 1982; Nelson & Frost, 1985). This network is considered important to contour integration (Li & Gilbert, 2002; Li et al., 2006; Polat et al., 1998), however top-down feedback also regulates behavior on this task (Li et al., 2008). Our results show that older observers can not perform inter-element linkages over the same distances as younger observers, which might reflect differences in anatomical or functional connectivity of long-range connections in striate cortex. Alterations to dendritic structure in primary visual cortex have been observed in elderly rhesus monkeys (Peters, Moss, & Sethares, 2001). Primate studies have also demonstrated delays of signal timing in cortical areas V1 and V2 (Wang, Zhou, Ma, & Leventhal, 2005), as well as elevations in spontaneous neural firing (Leventhal et al., 2003). One speculative possibility is that aberrant neural firing and timing delays in the aged brain might result in a decrease in neural synchronization that influences performance on tasks that require the grouping of perceptual information across space. Clearly, the decline or variation of neural machinery underpinning our measured behavioral differences requires further study. 
Acknowledgments
Research supported by Australian Research Council Discovery Project DP0877923 to author AMM. 
Commercial relationships: none. 
Corresponding author: Allison M. McKendrick. 
Email: allisonm@unimelb.edu.au. 
Address: Department of Optometry & Vision Sciences, The University of Melbourne, Parkville, Australia. 
References
Betts L. R. Sekuler A. B. Bennett P. J. (2007). The effects of aging on orientation discrimination. Vision Research, 47, 1769–1780. [PubMed] [CrossRef] [PubMed]
Betts L. R. Taylor C. P. Sekuler A. B. Bennett P. J. (2005). Aging reduces center-surround antagonism in visual motion processing. Neuron, 45, 361–366. [PubMed] [CrossRef] [PubMed]
Delahunt P. B. Hardy J. L. Werner J. S. (2008). The effect of senescence on orientation discrimination and mechanism tuning. Journal of Vision, 8, (3):5, 1–9, http://journalofvision.org/8/3/5/, doi:10.1167/8.3.5. [PubMed] [Article] [CrossRef] [PubMed]
Del Viva M. M. Agostini R. (2007). Visual spatial integration in the elderly. Investigative Ophthalmology & Visual Science, 48, 2940–2946. [PubMed] [Article] [CrossRef] [PubMed]
Field D. J. Hayes A. Hess R. F. (1993). Contour integration by the human visual system: Evidence for a local “association field”. Vision Research, 33, 173–193. [PubMed] [CrossRef] [PubMed]
Govenlock S. W. Taylor C. P. Sekuler A. B. Bennett P. J. (2009). The effect of aging on the orientational selectivity of the human visual system. Vision Research, 49, 164–172. [PubMed] [CrossRef] [PubMed]
Herbert A. M. Overbury O. Singh J. Faubert J. (2002). Aging and bilateral symmetry detection. Journals of Gerontology. Series B—Psychological Sciences and Social Sciences, 57, 241–245. [PubMed] [CrossRef]
Hess R. F. Field D. (1999). Integration of contours: New insights. Trends in Cognitive Sciences, 3, 480–486. [PubMed] [CrossRef] [PubMed]
Hess R. F. McIlhagga W. Field D. J. (1997). Contour integration in strabismic amblyopia: The sufficiency of an explanation based on positional uncertainty. Vision Research, 37, 3145–3161. [PubMed] [CrossRef] [PubMed]
Karas R. McKendrick A. M. (2009). Aging alters surround modulation of perceived contrast. Journal of Vision, 9, (5):11, 1–9, http://journalofvision.org/9/5/11/, doi:10.1167/9.5.11. [PubMed] [Article] [CrossRef] [PubMed]
Kovacs I. Julesz B. (1993). A closed curve is much more than an incomplete one: Effect of closure in figure-ground segmentation. Proceedings of the National Academy of Sciences of the United States of America, 90, 7495–7497. [PubMed] [Article] [CrossRef] [PubMed]
Kuai S. Yu C. (2006). Constant contour integration in peripheral vision for stimuli with good Gestalt properties. Journal of Vision, 6, (12):7, 1412–1420, http://journalofvision.org/6/12/7/, doi:10.1167/6.12.7. [PubMed] [Article] [CrossRef]
Leventhal A. G. Wang Y. Pu M. Zhou Y. Ma Y. (2003). GABA and its agonists improved visual cortical function in senescent monkeys. Science, 300, 812–815. [PubMed] [CrossRef] [PubMed]
Levi D. M. Yu C. Kuai S. Rislove E. (2007). Global contour processing in amblyopia. Vision Research, 47, 512–524. [PubMed] [Article] [CrossRef] [PubMed]
Li W. Gilbert C. D. (2002). Global contour saliency and local colinear interactions. Journal of Neurophysiology, 88, 2846–2856. [PubMed] [Article] [CrossRef] [PubMed]
Li W. Piech V. Gilbert C. D. (2006). Contour saliency in primary visual cortex. Neuron, 50, 951–962. [PubMed] [CrossRef] [PubMed]
Li W. Piech V. Gilbert C. D. (2008). Learning to link visual contours. Neuron, 57, 442–451. [PubMed] [Article] [CrossRef] [PubMed]
Loffler G. (2008). Perception of contours and shapes: Low and intermediate stage mechanisms. Vision Research, 48, 2106–2127. [PubMed] [CrossRef] [PubMed]
Mathes B. Fahle M. (2007). Closure facilitates contour integration. Vision Research, 47, 818–827. [PubMed] [CrossRef] [PubMed]
Mitchison G. J. Crick F. (1982). Long axons within the striate cortex: Their distribution, orientation, and patterns of connection. Proceedings of the National Academy of Sciences of the United States of America, 79, 3661–3665. [PubMed] [Article] [CrossRef] [PubMed]
Nelson J. I. Frost B. J. (1985). Intracortical facilitation among co-oriented, co-axially aligned simple cells in cat striate cortex. Experimental Brain Research, 61, 54–61. [PubMed] [CrossRef] [PubMed]
Owsley C. Sekuler R. Siemsen D. (1983). Contrast sensitivity throughout adulthood. Vision Research, 23, 689–699. [PubMed] [CrossRef] [PubMed]
Peters A. Moss M. B. Sethares C. (2001). The effects of aging on layer 1 of primary visual cortex in the Rhesus monkey. Cerebral Cortex, 11, 93–103. [PubMed] [CrossRef] [PubMed]
Polat U. Mizobe K. Pettet M. W. Kasamatsu T. Norcia A. M. (1998). Collinear stimuli regulate visual responses depending on cell's contrast threshold. Nature, 391, 580–584. [PubMed] [CrossRef] [PubMed]
Roudaia E. Bennett P. J. Sekuler A. B. (2008). The effect of aging on contour integration. Vision Research, 48, 2767–2774. [PubMed] [CrossRef] [PubMed]
Savage G. Haegerstrom-Portnoy G. Adams A. Hewlett S. (1993). Age changes in the optical density of human ocular media. Clinical Vision Science, 8, 97–108.
Sekuler R. Sekuler A. B. G., E. J. Williams, T. F. Beattie, B. L. Michel, J. P. Wilcock G. K. (2000). Visual perception and cognition. Oxford textbook of geriatric medicine. (pp. 874–880). New York: Oxford University Press.
Spear P. D. (1993). Neural bases of visual deficits during aging. Vision Research, 33, 2589–2609. [PubMed] [CrossRef] [PubMed]
Wang Y. (2001). Effects of aging on shape discrimination. Optometry and Vision Science, 78, 447–454. [PubMed] [CrossRef] [PubMed]
Wang Y. Zhou Y. Ma Y. Leventhal A. G. (2005). Degradation of signal timing in cortical areas V1 and V2 of senescent monkeys. Cerebral Cortex, 15, 403–408. [PubMed] [CrossRef] [PubMed]
Weale R. A. (1988). Age and the transmittance of the human crystalline lens. The Journal of Physiology, 395, 577–587. [PubMed] [CrossRef] [PubMed]
Wetherill G. Levitt H. (1965). Sequential estimation of points on a psychometric function. British Journal of Mathematical and Statistical Psychology, 18, 1–10. [PubMed] [CrossRef] [PubMed]
Winn B. Whitaker D. Elliott D. B. Phillips N. J. (1994). Factors affecting light-adapted pupil size in normal human subjects. Investigative Ophthalmology & Visual Science, 35, 1132–1137. [PubMed] [PubMed]
Figure 1
 
Examples of closed contour stimuli: A) a circular contour constructed of 10 elements; B) an elliptical contour constructed of 10 elements; C) a circular contour of 15 elements embedded in a noise field.
Figure 1
 
Examples of closed contour stimuli: A) a circular contour constructed of 10 elements; B) an elliptical contour constructed of 10 elements; C) a circular contour of 15 elements embedded in a noise field.
Figure 2
 
Mean (±95% confidence intervals of the mean) contrast thresholds for younger (blue) and older (red) participants.
Figure 2
 
Mean (±95% confidence intervals of the mean) contrast thresholds for younger (blue) and older (red) participants.
Figure 3
 
Mean (±95% confidence intervals of the mean) ellipse aspect ratio thresholds for varying numbers of elements comprising the contour for younger (blue) and older (red) participants. Results for stimuli presented within background noise are represented by the closed symbols (Panel A). Open symbols correspond to the condition without background noise (Panel B).
Figure 3
 
Mean (±95% confidence intervals of the mean) ellipse aspect ratio thresholds for varying numbers of elements comprising the contour for younger (blue) and older (red) participants. Results for stimuli presented within background noise are represented by the closed symbols (Panel A). Open symbols correspond to the condition without background noise (Panel B).
Figure 4
 
Mean (±95% confidence intervals of the mean) ellipse aspect ratio thresholds for varying amounts of Gabor patch orientation jitter for younger (blue) and older (red) participants. Panel (A), closed symbols, corresponds to the condition with background noise and panel (B) shows the results for stimuli presented on a blank background (open symbols). The data for sigma = 0 degrees are replotted from Figure 3, 15 elements.
Figure 4
 
Mean (±95% confidence intervals of the mean) ellipse aspect ratio thresholds for varying amounts of Gabor patch orientation jitter for younger (blue) and older (red) participants. Panel (A), closed symbols, corresponds to the condition with background noise and panel (B) shows the results for stimuli presented on a blank background (open symbols). The data for sigma = 0 degrees are replotted from Figure 3, 15 elements.
Figure 5
 
Mean (±95% confidence interval of the mean) threshold number of contour elements for shape discrimination for younger (blue) and older (red) participants, with (closed symbols) or without (open symbols) background noise. The right y-axis refers to the noise condition. The axes have been split to enable visualization of the magnitude of the significant interaction between noise condition and group.
Figure 5
 
Mean (±95% confidence interval of the mean) threshold number of contour elements for shape discrimination for younger (blue) and older (red) participants, with (closed symbols) or without (open symbols) background noise. The right y-axis refers to the noise condition. The axes have been split to enable visualization of the magnitude of the significant interaction between noise condition and group.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×