Visuo-spatial Cognition in Autism

1
Visuo-spatial Cognition in Autism
Lizzie Thomson1,2, Dagmara Annaz3 . Michael
Thomas3
2Department of Human Sciences, Brunel
University, UK 3Department of Psychology,
Birkbeck College, University of London, UK
Study 2 House Construction
Introduction
Autism Spectrum Disorder
PARTICIPANTS

• A rare neurodevelopmental disorder that affects
  approx. 60 in 10,000 children under age of eight
  (Wing Potter, 2002).
• General social deficit noted, with triad of
  deficits (Wing Gould, 1979) in socialisation,
  communication and imagination.
• Relative strength noted in piecemeal perception
  (e.g. Embedded figures test, Jolliffe
  Baron-Cohen, 1997).

• Overall participant group of 40.
• ASD participant group of 25.
• All 15 TD children completed all six house
  completions.

METHOD

• Physical integration of features in to an image.
• Six images of houses from a digital camera
  (Fujifilm, Finepix A202, taken by L. Thomson, May
  2004).
• Houses were matched to faces in featural/
  configural structure.
• Participants were timed while constructing blank
  house from selection of features (see example
  right).

Perception in Autism Low Level Processing.
Childrens Embedded Figures Test.
Block Design Test,

• Weak Central Coherence theory (e.g. Frith,
  2003)
• Inability to combine individual features in
  correct context and perceive in a configural
  manner.

• In this example of materials and set up, the top
  two features are foils and actual feature is on
  the bottom.

Perception and Construction of an Image.

• Strength of replication of perception in
• construction noted in autism.
• Participant E.C. drawing local-to-global
  (Mottron Belleville, 1993).
• Block Design test without segmentation (Shah
  Frith, 1993).

RESULTS
Face Processing in Autism and the Other Race
Effect.

• Well documented deficit in recognition of Other
  Race when compared to Same Race faces at a
  later date.
• Object processing route of OR faces results in
  superior performance in autism for later date
  recognition (Sasson Reznick, 2005)
• Holistic processing of SR and familiar OR
  faces, Tanaka, Kiefer Bukach (2004).

• Feature Selection (FS) and Configural Placement
  (CP) accuracy increases with age in ASD and TD
• (6p 0.042).
• No sig. difference between ASD and TD in FS or
  CP.

PARTICIPANTS

• 26 children aged 4- to 12-years (22 males)
  diagnosed as having an Autism Spectrum Disorder
  (ASD) and who met the CARS (Schopler, Reichler
  Rochen, 1993) criteria for high functioning
  autism.
• 15 children (11 males) aged 4- to 12-year were
  recruited as Typically Developing (TD)
  controls.
• Written informed consent was obtained from
  parents of all children.
• ALL 41 children completed various standardised
  tests to for cognitive and chronological matching
  for the two groups.

CONCLUSION

• Developmental trajectories of image construction
  accuracy is similar in ASD and TD groups.
• FS and CP are similar in ASD and TD.

For R2 rating white writing TD black writing
ASD. Each beside appropriate trajectory line.
6 (F(2,72) 3.800, p 0.042, Greenhouse-Geiser)
Study 3 Face Processing and The Other Race
Effect
STANDARDISED TESTS FOR ALL PARTICIPANTS

• ASD group CARS 35 (SD4.2).
• BPVS ASD 80 (SD12.2), TD 103 (SD6.6).
• Pattern Construction (BAS) ASD117 (SD31.5),
  TD 129 (SD21.2).
• Benton Faces SF Standardised score ASD 38
  (SD6.8), TD 42 (SD5.3).
• Copying (BAS) ASD 129 (SD18.6), TD 138
  (SD21.3). Note ceiling at 143.
• For standardised tests, the 2 groups were well
  matched for CA (p0.108). Cognitive profile
  shows a deficit in face recognition (BFRT
  Short-form, plt0.05), language (plt0.001) and
  copying (plt0.001). However, groups well matched
  on pattern construction (p0.205).

PARTICIPANTS

• Final ASD participant group was 14.
• All 15 children in TD group completed both
  tasks.
• No children were of Oriental Racial origin.

METHOD

• Discrimination of face images in half-tone (see
  right for example Caucasian
  study - top, and Asian - bottom).
• Items 1 6 direct matching of front view
  stimuli picture to one image from the
  multiple-choice display (choice of six images).
• Items 7 13 matching of a front view stimulus
  image with three images from three-quarter views
  (from six options).
• Items 14 16 matching of a front view image to
  three images from different lighting conditions
  (again, from six options).

Due to CA matching, all standardised test
group- level comparisons used ind. t-test.
Study 1 Low-level Contour Integration
RESULTS

• Main interest is in difference between
  performance of the 2 groups on the 2 tasks.

• No sig. difference between the face recognition
  rates in each test.
• No sig. difference between the groups and
  overall percentage accuracy.
• Only Asian test showed sig. improvement with CA
  (7p lt0.05)

PARTICIPANTS

• 23 of the ASD group participated.
• All 15 children in the TD group completed both
  contour-detection tasks.

METHOD

• Task Previously used to demonstrate gradual
  integrative nature of perception in typical
  development (Kovács, 2000).

• Two card-test versions of a contour-detection
  task were used. One grey-scale using Gabor
  contour signals (see example) the other using
  colour to indicate the circles.

CONCLUSION

• One grey-scale using Gabor contour signals (see
  example) the other using colour to indicate the
  circles.
• Psychophysical indication of one circle per card
  until inability of circle detection is noted
  (threshold).

• All participants improved Asian, but not
  Caucasian face detection skills with
  chronological age.
• No significant difference between the groups in
  face recognition.

Circle to detect
For R2 rating white writing TD black writing
ASD. Each beside appropriate trajectory line.
7t 2.138, p lt0.05
Summary and Discussion

• In a low-level perceptual task, children with
  ASD demonstrated a preference for featural
  processing and development of contour integration
  had a different relationship with CA for the 2
  groups.
• However, a high-order image reproduction showed
  similar developmental patterns in ASD and TD
  groups (according to CA) in featural and
  configural placement
• Face processing tasks did not demonstrate either
  a SR/OR bias or a difference in recognition
  accuracy percentages between the 2 groups against
  CA.
• However, the cognitive profile demonstrated that
  the children in the ASD group were not
  particularly high functioning and a sig.
  difference noted on SF stand. score on BFRT.

RESULTS

• Expected developmental pattern was a reduction
  in threshold level with age.
• ALL participants demonstrated a significant
  increase in contour detection with increasing
  chronological age (CA), 3p0.008.
• Similar overall contour detection rates were
  noted in both groups (4p0.743).
• However, participants with ASD did not
  demonstrate a typical developmental trajectory
  as they showed increasing divergence from normal
  trajectory with CA (5p0.016).

OVERALL CONCLUSIONS

• Visuo-spatial skills increase with CA,
  regardless of whether the individual is TD or
• has an ASD.
• Only low-level perception in autism demonstrated
  a non-typical developmental pattern with CA.

References
For R2 rating white writing TD black writing
ASD. Each beside appropriate trajectory line.

• Frith, U. (2003). Autism Explaining the Enigma.
  (2nd ed.) Oxford Blackwell.
• Jolliffe, T. Baron-Cohen, S. (1997). Journal
  of Child Psychology and Psychiatry, 38, 527
  534.
• Kovács, I. (2000). Vision Research, 40, 1301
  1310.
• Mottron, L. Belleville, S. (1993). Brain and
  Cognition, 23, 279 - 309.
• Sasson, N. J. Reznick, J. S. (2005). Biennial
  Meeting of The S.R.C.D. Atlanta, USA.
• Shah, A. Frith, U. (1993). Journal of Child
  Psychiatry and Psychology, 34, 1351 1364.
• Tanaka, J. W., Kiefer, M. Bukach, C. M.
  (2004). Cognition, 93, B1 - B9.
• Wing, L. Gould, J. (1979). Journal of Autism
  and Developmental Disorders, 9, 11 29.
• Wing, L. Potter, D. (2002). Mental Retardation
  and Developmental Disabilities, 8, 151 161.

3F(1,34) 8.02, p0.008 4F(1,34) 0.110,
p0.743 5F(1,34) 6.43, p0.016
CONCLUSION

• All participants demonstrated an increased in
  contour-detection ability with CA,
• regardless of participant grouping.
• There was no significant difference in the
  overall contour detection ability between the
• participant groups.
• However, children with ASD demonstrated an
  increasing divergence from the expected
• developmental trajectory with CA. This suggest
  that the relationship between contour-
• detection and CA is different in ASD when
  compared to TD.

With special thanks to Elisabeth Cook and ABC
after school club, Ruislip Gardens School,
Hillingdon. Hillingdon Manor Lower and Middle
School, Hillingdon, Margaret Thomson, Rachael
Gabriel and all the children who participated.
1Contact Detailselizabeth.thomson_at_brunel.ac.uk Th
is research was conducted in partial fulfillment
for the award of M.Sc Cognitive Neuropsychology.
UCL, University of London, UK.