Computational modelling of visual disorders

Visual disorders can occur due changes at any stage of visual processing.
Glaucoma affects the earliest stage in the retina, and leads to blindspots. In
contrast, patients with amblyopia can have perfectly healthy eyes, and it is
thought of as a cortical disorder. There is a long standing debate about how
the brain is altered in these disorders [1]. Is there neuroplasticity to
compensate for these impairments? If so, at what stage of visual processing do
these adaptations occur?

Functional magnetic resonance imaging (fMRI) allows us to monitor activity in
the visual cortex noninvasively. Computational models provide an explanation
for this activity, i.e., how the visual information is being processed. These
models can be used to compare patients to controls exploring how visual field
maps and receptive fields are altered [2]. More specifically we will use
sophisticated "population receptive field" [3] models which predict how
specific populations of neurons will respond to the spatial extent of stimuli;
and the "neural contrast sensitivity function" [4] which predicts the
sensitivity of populations of neurons to different stimuli. Together these
models will constitute a *computational* perspective to complement the existing
anatomical and behavioural understanding of these disorders, potentially
providing novel methods for tracking progression and treatment.


[1] Wandell, Brian A., and Stelios M. Smirnakis. "Plasticity and stability of
visual field maps in adult primary visual cortex." Nature Reviews Neuroscience
10.12 (2009): 873-884.
[2] Dumoulin, Serge O., and Tomas Knapen. "How visual cortical organization is
altered by ophthalmologic and neurologic disorders." Annual
Review of Vision Science 4 (2018): 357-379
[3] Aqil, M., Knapen, T., & Dumoulin, S. O. (2021). Divisive normalization
unifies disparate response signatures throughout the human visual hierarchy.
Proceedings of the National Academy of Sciences, 118(46), e2108713118.
[4] Roelofzen, C., Van Dijk, J. A., De Jong, M. C., & Dumoulin, S. O. (2020).
Measuring contrast sensitivity functions in human visual cortex. Journal of
Vision, 20(11), 1379-1379.

Primary Objective: To explore the computations which occur in the brain which
occur in visual disorders and determine how these differ from healthy controls.
Specifically, we expect that the overall sensitivity (area under the curve)
which reflects the responsiveness of a brain region to images of different
contrasts, will be reduced in patients compared to controls.
Secondary Objective(s): To compare other model parameters (i.e., pRF size) (and
their differences) with basic visual processing (i.e., acuity, visual field
deficits and contrast sensitivity).

The study will be an observational, cross-sectional study comparing healthy
controls and patients (amblyopia, glaucoma). Contact will occur through two
routes. [1] Physicians: patients will be given study information by the
relevant physicians, if they are interested in taking part they will then
contact the project leader (Marcus Daghlian). [2] Flyers, online advertisement
and word of mouth will be used to recruit additional participants. If
interested they will contact the project leader (by email), who will provide
them with the full information sheet (E1 E2). The project leader will then call
the participants, ensure that they understand and have read the information
sheet; perform screening and obtain informed consent. Participants will be
contacted by the project leader to visit the study centre for two separate
sessions (screening and informed consent will be recollected on both sessions).
Session one will consist of anatomical MRI scans; functional scanning for the
first eye (presenting participants with low level visual stimuli, while one eye
is patched) and basic psychophysical tests to assess the level of visual
function (again for the first eye). The second session will be identical to the
first, but the second eye will be patched, and anatomical data will not be
reacquired. Participants will lie on the MR (Philips 7T scanner) table while
the scan takes place (maximum duration 1hr 30mins), either resting (during
anatomical scans) or viewing stimuli (during functional scans). Visual stimuli
will consist of synthetic images (e.g., textures and patterns) or natural
images (i.e., scenes, objects); these images will not contain any content which
could be offensive or stressful (i.e., pornography, violence, emotive scenes
etc.). During the functional scans, participants will perform a task;
maintaining fixation (looking at the centre of the screen) and pressing a
button in response to a change in the stimuli.

There are no risks associated with participation in this study, provided the
exclusion criteria for fMRI are taken into account. The burden of participation
is mainly restricted to the time involved in the measurements; in each of the
two sessions there will be a maximum of 1 hour of psychophysical assessment
and a maximum of 1 hour and 30 minutes in the scanner. Scan sessions will take
place over separate days, unless the participants request otherwise to reduce
their travel time. In the latter case there will be a minimum of 2 hours
between sessions to ensure participants are well rested. Functional MRI is a
non-invasive technique, and there is no need for special preparation for the
subject. The psychophysical tests involve button presses in response to basic
(non-emotive, non-explicit) images. For both fMRI and psychophysical testing,
participants will have one eye patched, this may be mildly uncomfortable. The
patch will only be applied while the participants are seated / lying,
immediately before the tasks / scans respectively, and will be removed as soon
as the task / scans have finished. .

The MRI data are used for research purposes only. However, severe abnormalities
may be noticed, in which case a specialist (radiologist) may be asked for
advice, upon decision of the research team. If the specialist confirms that
medical treatment is indicated, then the subject participant*s general
physician will be notified. If the abnormalities are classified as untreatable
by the specialist, no information will be passed on to the general physician.
The volunteer is informed of this possibility prior to participating in the
experiment, and signs a separate paragraph on the informed consent form to
signal their consent with this procedure. If they do not consent to this they
will not be scanned.

The participants of this study will have no direct benefits from participating,
aside from financial compensation for time.