ALSPAC provides an ideal opportunity for a pilot study to explore the extent to which the newly-designed "contour card" test indicates the integrity of any visuoperceptual functions in children aged 11-13 years. ALSPAC is an ongoing birth cohort study with a large amount of prospectively-collected data on visual functions and other abilities, in a contemporary sample of people who are now in their late teens. The contour cards were used in one of the ALSPAC clinics when the children were aged 11 years. An estimate of visuoperceptual skills in the children has been obtained by asking their mothers a battery of questions about their child's visuoperceptual abilities (based on Houliston et al., 1999). Both the raw visuoperceptual question responses, and the results of a factor analysis of these data, are being prepared for publication. We propose to compare the ALSPAC cohort members' performance in the contour card test at 11 with their visuoperceptual skills (as elicited in these questions), and their reading abilities, as theory and previous research predict that these might be linked with performance in the contour test. We would also include basic demographic data such as age and sex, and relevant eye problems such as strabismus and amblyopia. Outputs would be an MRes thesis to University of Bristol and if appropriate, a paper in a peer-reviewed journal.

Background

One of the oldest unsolved problems in vision science is how the visual system groups spatially separate elements of a display to form a coherent whole (May and Hess 2007). In vision science this is termed perceptual grouping, which refers to our ability to extract visual information from lower-level features to obtain perception of a higher-level structure. Psychophysical research has attempted to delineate principles by which local orientations and motions are combined across space to facilitate the detection of simple spatial contours. The perceived global structure of a scene depends heavily on the configuration and properties of its constituent local visual features (Ledgeway et al., 2005). Kovacs et al., 1999 developed a test of perceptual grouping, using "contour integration", that is suitable for use with paediatric patients in a clinical setting. This was motivated by research (1996) which found the detectability of a closed path (contour) defined by Gabor elements is significantly degraded in amblyopic eyes. These studies found that performance was degraded in strabismic, but not anisometropic amblyopia. Cells in the early parts of visual pathway have small receptive fields that are well tuned for orientation, direction of motion, and disparity (Norcia et al., 2005). Contours defined by Gabor patches cannot be detected by large filter mechanisms operating on the scale of the contour itself. Such elements force the observer to detect the contour on the basis of long-range comparisons between local orientation measurements at the defined spatial scale. The contour test has been devised to test perceptual grouping and has been shown to detect differences in people without ocular problems vs. those with amblyopia, as the above references mention. However, it has not been demonstrated whether the contour test can be useful as a measure of higher visual functions in children. We are interested to test children with problems in integrating different aspects of visual information (ie visuoperceptual problems), specifically those with "simultanagnosia". "Simultanagnosia" is a visuo-perceptual difficultly linked to object recognition in a crowded scene. This is thought to involve difficulties in using perceptual grouping to see the bigger picture made up from a number of component elements, thus "simultanagnosia" may be associated with poor performance in the contour test.

Details of the Contour detection card test

The cards consist of a closed chain of Gabor signals that roughly model the receptive field properties of orientation-selective simple cells in V1 (Kovacs et al., 1999). Contours defined by Gabor patches force the observer to detect the contour on the basis of long-range comparisons (Pennefather et al., 1998) for orientation differences of up to +/- 60(masculine ordinal indicator) amongst elements, which implies an association field and the path is reliably identified (Field et al., 1993; Loffer, 2008). Normal observers' performance on contour task starts to degrade at contrasts below 7-10% whilst amblyopes had thresholds outside of the 95% confidence interval range of normal observers (Kovacs et al., 2000). Contour integration deficits thus constitute an aspect of visual function that is disrupted by abnormal visual experience (Pennefather et al., 1999). (See below for an example of contour detection card).

The parameter (delta) is the ratio of background element spacing to contour element spacing. Variation of the parameter allows one to isolate first-order and second-order integration mechanisms, since the detection of the contour at values of (delta) less than 1 is impossible using first-order cues alone. The advantage of using (delta) parameter is that one can directly study the efficiency of long-range interactions that contribute to the integration of spatially distributed objects. By changing (delta), only the signal-to-noise ratio changes, while the shape of the contour, the global and local curvatures, the number of contour elements, the length of the contour, the spacing along the contour and the eccentricity of the elements are all kept constant. Contour integration deficits were found predominantly when the task required second-order information (ratio drops below 1) (Kovacs et al., 1999).

Few studies have been conducted to explore to what extent the contour test reflects day-to-day performance in particular tasks. However, adults with reading difficulties (diagnosed with dyslexia) have been reported to perform less well in the contour test than age-matched adults without reading problems (Simmers et al) .

Simultanagnosia

Simultanagnosia is a visual deficit described as a relative inability to comprehend more than one visual element in a scene and hence picture the whole. It has been described after prenatal damage to the brain, or after later-acquired parietal lobe injuries. We would like to test for covariation between results in the test of 'perceptual grouping' (the contour test) in children aged 11 and the reports of childrens' mothers on their abilities in the questions designed to elicit symptoms of visuoperceptual difficulties, including simultanagnosia. The questions asked in the ALSPAC questionnaire were originally devised to aid history-taking in patients with hydrocephalus, who often have vision and visuoperceptual problems. Questions about visuoperceptual problems have not been widely used in samples of normally developing children. The ALSPAC mothers' responses to the questions have been analysed looking for common underlying themes, using principal components analysis (PCA) and the data suggest that the mothers' responses cluster around 3 "factors": Factor 1 correlates with questions about seeing objects surrounded by clutter (similar to simultanagnosia as described above); Factor 2 correlates with visuomotor problems (difficulties making judgements about positions of self or objects within 3D space), and Factor 3 with questions about recognition of faces. The first two are thought to be mediated primarily by a part of the brain known as "the dorsal pathway" and the latter by another area called "the ventral pathway". These hypothesized functional sub-divisions within the brain are based on experimental evidence and on observations in clinical literature: the dorsal pathway is thought to carry information about where objects are in space, whilst the ventral pathway is more responsible for fine detail and recognition.

Objective

We would like to test the childrens' raw visuoperceptual scores as reported by the mothers, and 3 PCA-derived factors of visuoperceptual skills, in comparison with the contour test results. We will take account of other potentially important variables such as visual acuity (at appropriate age) and factors known to impair performance in the contour test eg strabismus and amblyopia. We would also like to look at a measure of reading ability with respect to the contour results, to estimate whether the contour test will usefully predict reading difficulties in this cohort. We hypothesize that the contour test results will show stronger associations with Factor 1 than with Factors 2 or 3. We also hypothesize that some children with reading difficulties (ie reading worse than that predicted by their IQ) will perform less well in the contour test than will children with reading skills commensurate with their IQ.

We therefore request access to these specific variables:

* Child's sex

* Mother's highest educational attainment

* Parent's socio-economic status (SES from pregnancy questionnaires)

* Child's reading ability at 13 (TOWRE test at tf2)

* Childs IQ (WISC at F8 clinic)

* Child's visual status (strabismus and/or amblyopia assessed at F7 clinic)

* Child's contour test results (F11 clinic)

* Child's visuoperceptual skills (questions asked of mum in 13-yr child-based questionnaire).

The analytic strategy will start with tests of correlation, either parametric or non-parametric depending on the distribution of the data. We will then use Logistic regression and/or Generalised Linear Modelling (GLM) to obtain estimates for the associations between the contour test results and either reading or visuoperceptual skills, adjusted for relevant potential confounders.

References

Field, D.J., Hayes, A., Hess, R.F. (1993) Contour integration by human visual system: evidence for a "local association field" Vision Res, 33, 173-93

Houliston, M.J., Taguri, A.H., Dutton, G.N., Hajivssiliou, C., Young, D.G. (1999) Evidence of cognitive visual problems in children with hydrocephalus: a structured clinical history-taking strategy. Dev Med Neurol, 41, 298-306

Kovacs, I., Kozma, P., Feher, A., Benedek, G. (1999) Late maturation of visual spatial integration in humans. Proc Natl Acad Sci USA, 96, 12204-9

Ledgeway, T., Hess, R.F., and Geisler, W.S. (2005). Grouping local orientation and direction signals to extract spatial contours. Empirical tests of "association field" models of contour integration. Vision Research, 45, 2511-2522

Loffler, G. (2008) 'Perception of contours and shapes: Low and intermediate stage mechanisms', Vision Research, 48, 2106-27

May, K.A. And Hess, R.F. (2007). Dynamics of snakes and ladders. Journal of Vision, 7, 1-9

Norcia, A.M., Sampath, V., Hou, C. (2005). Experience-expectant development of contour integration mechanisms in human visual cortex. Journal of Vision. 5, 116-130

Pennefather, P.M., Chandna, A., Kovacs, I., Polat, U., and Norcia, A.M., (1999) Contour detection threshold: repeatability and learning with 'contour cards'. Spatial Vision 12, 257-266

Simmer and Bex. Deficits in Visual Contour Integration in Dyslexia. IOVS 2001 42: 2737 - 2742.