Virtual reality & therapeutic applications in autism

Dr David Simmons

therapeutic applications of VR in autisim

• phobia

• social skills

  • (autism controversial)

• anxiety

specific phobia

• extreme or irrational fear of an object or situation 

• common in autism 

  • 30-64% of autistic children

VR for specific phobia rationale

• graded exposure combined w/CBT known to effective 

• usually progress from imaginal exposure to real-life

• but some people struggle w/imaginal exposure & real life exposure is hard to control 

why use VR for specific phobia

• allows simulations of RW to be created

• newly learned coping skills can be rehearsed & reinforced in safe & controlled way

• facilitates a more gradual exposure

• a good alternative 

Maskey et al (2019) - blue room

• Looking into specific phobias 

• single blind randomised controlled trial

  • VR treatment vs usual care

• built on previous proof of concept

• outcome measures of comparision of

  • initial vigette

  • produced by blind assessor

  • rated by expert panel

    • to see how p responds

Maskey et al (2019): Methods

• Careful Onboarding procedures

  • bringing P into enviroment 

• Home visit with video of procedure

• Relaxation before exposure

• CAVE VR (No headset) 

  • 4 m3

  • Therapist sits next to child to deliver CBT

• Broad range of measures

• Children seemed to enjoy the procedure

• Low dropout rates

Maskey et al (2019): results 

• gradation allowed P to experience levels of mastery in managing their anxiety and to repeatedly practice 

• most children’s rating of their confidence at tackling their goal situation increased

• an increase reflected in parent ratings

• Data completeness for the main outcome measure of target behaviour rating was excellent

• a success rate of 38% with definite positive change on blinded, independent rating, with a large effect size, is at least comparable with other interventions for anxiety

evaluation of Maskey et al (2019)

• Innovative use of VR in a very specific context

• Cost of individual adaptation

• Potential bias from non-blinded parents

• Unclear why works well for some but not others

• Comparison vs. “treatment as usual” but not “VR vs. non-VR”

Autism

• common form of ND which directly effect 1-2% of pop

• characterised as having difficulties w/social interactions

  • recasted as differences rather than deficits

• social skills training used to be seen as useful but now seen as highly controversial

• tend to have bias towards local processing due to weak central coherence

Autism & social skills rationale

• If role-playing social situations is thought to be a good thing, alternatives are:

  • Reading and imagining

  • Watching videos

  • Live role play w/actors

• VR offers another alternative which is

  • Safe

  • Reproducible

  • Customisable

Parsons et al (2004)

• 36 participants

• 12 Autistic, 12 controls matched on Verbal IQ, 12 on Non-verbal (”Performance”) IQ

• “Desktop” VR (Cheap, Accessible

  • Less problems of “cybersickness”

• PoV display, controlled by a joystick

  • “Training Environment”

  • “Virtual Café

Parsons et al (2004) results 

• learned to use the equipment quickly 

• showed significant improvements in perfomance

•  VR technologies are potentially trasnformative for education

• tendency to walk into objects/people could not be explained by executive dysfunction or a general motor difficulty → might be a sign that understanding personal space is impaired in autism

• structuring of tasks within the environments provides extra support for people with weaker executive abilities

Parson et al (2004) evaluation

• 20 year old

  • outdated langauge to describe autism

  • outdated attitudes towards social skill training

• early influential study

• intresting that there wasnt a issue as it wasnt similar to the experience 

anxiety 

• A mental state characterized by apprehension, avoidance, and cautiousness regarding potential threats, dangers and negative events

• Psychobiologically distinct from fear

  • Response to imminent danger

• Can be useful

  • Assisting w/uncertain threats

  • Mitigation/avoidance of future events

• Can be maladaptive

  • Negative impact on mental well-being

  • Particular problem for autistic people (27-50%)

manipulating anxiety rationale  

• evidence that sensory features of autism vary w/anxitey

• evidence is largely from questionnaires & self-reports

• useful to manipulate anxiety in controlled, reproducible & safe environment

Millington (2023)

• 48 participants

• Completed autism trait and anxiety questionnaires

• State Anxiety manipulated using custom designed VEs

• Performance in sensory task measured after each manipulation

• HMD (HP Reverb Omni) with associated pupil & heart rate measures

• performed at Glasgow ARC/XR lab and Engineering laboratories

Millington (2023) findings

• state anxiety varied significantly between virtual environments

• task did not reveal any significant effects of anxiety or autistic traits

Millington (2023) evaluation

• Effects on perceptual task almost significant

• Not many diagnosed autistic participants

• Some technical issues

• Face Makeup smudging sensors

• Consistency in auditory stimuli

• Pupil size varies with display brightness

• Some participants found the forest environment scary!

• Anxiety maybe increased with unfamiliar tech

Parson (2016) Authenticity in VR for assessment & learning

• VR is a promising tool for learning & assessment, especially for individuals with autism.

• Authenticity = key to ecological validity (VR behaviour reflects real life).

• Designers must prioritise essential authenticity dimensions (visual, functional, contextual).

• Functional fidelity often matters more than visual realism; test VR tasks against real-world performance.

authenticity (Parsons, 2016)

• how closely the VR scenario maps into real-world tasks/contexts

• more authentic a VR system the stronger its validity for assessment 

issues of authenticity (Parsons, 2016)

• more authentic --> more it costs, technical complexity & usability constraints

• over-realistic can overwhelm users or reduce usabilityOver-realistic