30366
The Use of Immersive Virtual Reality to Teach Safety Skills to Children with Autism Spectrum Disorder

Poster Presentation
Friday, May 3, 2019: 10:00 AM-1:30 PM
Room: 710 (Palais des congres de Montreal)
C. Miyake1, E. Hong1,2 and D. Dixon1, (1)Center for Autism and Related Disorders, Woodland Hills, CA, (2)Louisiana State University, Baton Rouge, CA
Background: Individuals with autism spectrum disorder (ASD) experience impairments across a wide range of skill domains, including safety skills. Teaching safety skills to individuals with ASD is critical, given the potentially dangerous and fatal consequences. Although several methods (e.g., video training, mock simulations, natural environment training) have been used to teach street-crossing skills to children with ASD, these methods have largely been ineffective in teaching generalization of skills to real-life street-crossing scenarios. While natural environment training poses the lowest barriers to teaching generalization, it often proves to be the most difficult to implement due to safety and logistical challenges (i.e., the ability to present both safe and unsafe conditions in an efficient manner). Virtual reality (VR) technology may be a solution to teaching safety skills to individuals with ASD, such that VR environments provide immersive, realistic scenarios in a safe, efficient, and controlled manner.

Objectives: The objectives of this study were to determine if (1) safety skills could be taught in the VR environment and (2) safety skills learned in the VR environment could be generalized to the natural environment.

Methods: The current study used a multiple baseline design to teach three children with ASD (5-8 years old) street-crossing skills. All natural environment pre-treatment, post-treatment, and probe sessions were conducted in-vivo, in uncontrolled traffic areas. All immersive VR probe and training sessions were conducted using the Oculus Rift headset and sensors. The training sessions were conducted in a 10’x12’ therapy room equipped with a table that contained the Oculus Rift headset and sensors. The primary investigator used a laptop connected to the Oculus Rift to monitor what the participant was seeing in the VR environment, in order to present the questions and to score the participants’ responses. The video clips were hosted on YouTube and accessed through the Oculus and Steam software.

Following completion of the pre-treatment baseline and probe sessions, participants began immersive VR training using short, 10-second video clips of safe (i.e., no car present) or unsafe (i.e., car present) conditions. Depending on the participant’s progress after VR training using the short clips, they either proceeded to training using 10-second video clips with added audio distraction (e.g., dogs barking, lawnmowers, etc.) or directly to long, 3-5-minute uncut clips of a street.

Results: During the pre-treatment baseline sessions, all three participants averaged below 50% on their identification of safe and unsafe conditions in the natural environment. Following treatment, all three participants scored 100% in the post-treatment probe sessions. Participants 2 and 3 only needed the short clip and long clip training video types to reach mastery criterion in the natural environment, while participant 1 needed training in all three training video types.

Conclusions: The current findings suggest that immersive VR training may be a safe and viable method of teaching safety skills to individuals with ASD.