Monday, 1 April 2013

What proportion of adults who are dyslexic would respond to visual intervention?

What proportion of adults who are dyslexic would respond to visual intervention?
Data from Bradford University UK

We are about to hear from the Oxford-Kobe symposium which will include reports on the visual processing aspects of dyslexia as well as phonological processing.
In the last few years there have been few accepted studies in the area of visual interventions.
One report that came out was the outcomes by a study by Jim Gilchrist and others at Bradford University School of Optometry and Vision Science in the UK...
Gilchrist and his colleagues have looked at visual stress problems experienced by Dyslexic undergraduates at Bradford University.

There were 52 dyslexic students referred to the clinic at Bradford Department of Optometry and vision Science. They were taken through specific protocols.
The idea was to identify any visual dysfunctions/ stress that could be an issue for the students and contribute to their problems with text.

As Vision Scientists, they wished to distinguish between reading related visual stress
Originating in Occulomotor / optometric / ophthalmic issues.
Visual stress originating in visual processing problems associated with the pattern or the spatial characteristics of the task, From now on referred to as PRVS.

Occulomotor / optometric / ophthalmic .dysfunction

Did they experience visual stress associated with an optometric dysfunction?

59% of the students had no evidence of any accommodation or vergence problems.
Vergence problems are problems which can be measured as possible faults in the way the muscles maintain stable binocular vision. Usually the intervention is by the use of prisms or exercises optometric or orthoptic Vision therapy.
This component was looking at general uncorrected optometric dysfunction. The vergence dysfunction would have, I assume have been considered at near.

As you can see around 60% showed no evidence of problems with any evidence from this study as possibly having any origin in this area.

 Did they exhibit indicators of Pattern Related Visual stress?  (The PRVS study did not I believe include a reading speed/RAN measure which I would argue is a key indicator.)

Two thirds showed evidence of Pattern related Visual stress...  Colour mediated.
The next question is whether those experiencing PRVS might also benefit from optometric intervention.

Of those who experienced PRVS, 31% also experienced visual stress originating in accommodation (uncorrected focussing problems) and vergence problems. That is 20% of the students.
69% of the students experiencing PRVS had no issues at all which would benefit from optometric intervention.  That is 45% of all the referred students.
A further 20% of the students showed no signs of any issues associated with optometric intervention.

15 % of the referred students showed any evidence of any dysfunction, either PRVS or optometric in origin.  This suggests that their phonological difficulties were totally phonological in origin with no visual component.

The statistics above suggest a moderate relationship between accommodation and vergence dysfunction. This would be expected in terms of the tendency to suppress an eye which is giving a weaker set of visual data from poor focussing.

Importantly it suggests that the PRVS dysfunction is NOT an ophthalmic originating issue but is separate and by implication should be subject to separate intervention. There is a post on the blog which covers two case studies which attempt to quantify the benefits from two separate but parallel interventions .

This would be in agreement with a link to the spatial characteristics of the retina and visual system.
Foveal diameter.
Crowding effects associated with the centre –surround system for edge detection. The diameter of these units being associated with the ratio of red to green sensitive cells which varies from person to person.
Crowding effects associated with the variation in cone cell diameter between people. Some would suggest an inverse relationship between the central foveal cone cells and the diameter of the fovea itself.
Bigger fovea…

 Probably big visual attention span more characters parallel processed at a time
Reduced foveal crowding
Increased perceptual span... more characters per fixation.
Longer saccades

Possible weaknesses in the study
This study can be looked at critically in terms of what definitions and protocols were applied

 For example…
Were the students seen by the department representative of the whole body of dyslexic students at Bradford?
What was the actual definition of PRVS?
It can be argued that the symptoms of PRVS are not necessarily perceived by the ‘person’ it could simply put extra demand on the central executive reducing phonological output in terms of speed and ‘quality’ (prosody/intonation)

What I do know is that the academics, that ran this study, are of the highest calibre and as such the data should not be ignored.
What was reported in a lecture by Jim Gilchrist, was also that many of those who did not get identified as having PRVS, by the selection criteria, nevertheless did respond positively to the use of colour chosen using the Intuitive overlay forced choice selection process. Although this process itself is from a very limited palette.  But no one could understand why.

The idea that several of those not identified as having (the symptoms) of visual stress benefitted from the use of colour can be used to question whether the symptoms of PRVS should be the criteria which limit access to interventions.

If we look instead at the work of Valdois and Facoetti on visual span and crowding and hypothesise that any restriction on visual attention span will decrease the phonological output.
Whether that restriction be from
Focussing problems
Edge detection processing speed
Vergence difficulties... binocular problems
Reducing working pigment density in the cone cells.

Increasing the number of letters processed between saccades

Restriction in visual attention span and reduced edge detection processing will give rise to the following changes which will further limit both visual and phonological performance probably the basis of the reading stamina difficulties virtually universal amongst dyslexic adults.
decrease the lengths of the saccades
Increase the occulomotor muscle tone, decreasing ocular motility.
increase the demand on working memory ( central executive)
increase the difficulties of blending phonemes
slow down the development of automaticity
Reduce the cumulative total reading experience.

Binocular dysfunction
A way of looking at the effects of binocular dysfunction on visual processing speed would be to compare rapid automatic naming speeds under the following three conditions.

Monocular right eye … left eye occluded
Monocular right eye occluded.

When I have done this with dyslexic individuals, it is very rare that binocular reading gives rise to greater reading speeds than both monocular.
Usually either the right eye or the left eye gives the greatest reading speed.
Binocularly the reading speed is more often the same as the left or right monocular speed. Sometimes but not often it is slower than either.

When the computer screen has been optimised, then usually this is faster than either eye monocularly.

This suggests that the eyes are then acting synergistically.

Binocularity and visual processing speed.
The binocular instability hypothesis that seems central to the magnocellular deficit hypothesis, suggests that by occluding one eye the reading should be faster, more effective.
But very few dyslexic undergraduates get rid of their ‘visual stresses by occlusion.  Those who can get rid of the visual stress by occlusion, have nearly always been using this already as a strategy to extend their reading stamina and decrease the discomfort.
The dyslexic undergraduates (adults) who report instability of the text usually report it still when viewing text monocularly. The instability has to be then associated with spatial and temporal visual data being collected by the viewing eye. This implies then that the dynamics of image movement on the retina during fixation is important in the edge detection process. When I was participating in a pilot study of crowding, where the image was projected for 50ms time periods I was very aware of ‘seeing’ several images during the exposure as my eye moved extensively within the 50ms period. During a fixation visual data is being collected for up to 200 mS.
When you look at the binocular eyetraces for these people there is usually clear evidence that the visual system is already, subliminally suppressing the data from one eye anyway.
You can see this in some of the other posts for example.

You can calculate the optimal screen settings for a person, monocularly and then apply the settings binocularly. This implies that what is being measured is associated with the biology of a single eye, rather than binocular aspects of the system.
The way the eye movements of both eyes are affected and the synergistic reading performance benefit when reading binocularly could be the basis of further conjecture about the system that is operating.

As a final point; when I have looked at reading performance of fluent readers, the binocular reading performance exceeds monocular performance. In addition when reading English (left to right) the left eye appears to be landing in front of the right eye. I have only ever looked at a fluent bilingual person once. That person was bilingual in Urdu (right to left) and English.  Whichever text was being read the left eye was always in front. This concurs with Gadi Geiger’s work at MIT on visual attention from a ‘pre-computer age. I had a great discussion in Boston, Mas. on this issue ‘earlier this century’.

Fixation disparity
Work with Arnold Wilkins et al, looking at the reading performance of year 7 (6 in the USA) and measuring accommodation, orthoptic and PRVS’ showed only one correlate with default reading performance across the group, fixation disparity.
The fixation disparity, it would seem needs to exist to be a fluent reader!  If there is no disparity, then the reading performance is lower, if it is too high the reading performance is lower. There wo0uld appear to be an optimal fixation disparity for a person.
This would make sense in terms of the way many dyslexic people, but not all are gifted graphically. There seems to be an excess of dyslexic students studying architecture.
It may also be one of the reasons behind why converting text into graphics, by such means as Mind mapping, is reported as being a particularly useful study aid for dyslexic students.
What the use of a binocular eyetracker tells us though is that the amount of binocular disparity employed by a person’s visual system is to some extent task responsive.  Again this would concur with Geiger’s research. The disparity in fluent readers is directional depending on the brains identification of the reading direction needed by the language.

No comments:

Post a Comment