16040
A Step Towards Anxiety-Sensitive Virtual Reality Based Social Communication Platform: Implication on Physiology for Children with Autism
Social stories can be an effective way in addressing the core social communication deficits in children with autism spectrum disorder (ASD). Effective skill-learning is possible when an individual learns from his comfort zone with diminished level of anxiety. Anxiety disorders in children with ASD may be considered both as possible cause and consequences of their social communication deficit. Providing a platform that monitors the anxiety of these children may be beneficial. Specifically, children with ASD have difficulty in explicitly expressing their affective states, which places limits on conventional observational techniques. The anxiety-sensitive platform can be used to overcome such limitations using physiological markers of one’s affective states. Among the available technology-assisted techniques we chose Virtual Reality (VR) for developing the social communication platform along with physiology as a tool for prediction of one’s affective state. In India, with prevalence rate of 1 in 250 children with autism, and scarce trained resources, research on technology-assisted systems for these children is necessary.
Objectives:
The presented work seeks to address the deficits in social communication of children with ASD by (i) developing a VR-based social communication task module considering language and subtle social aspects specific to Indian context, which are valuable contributors to social communication (ii) interfacing the VR-based system with physiological data acquisition module, and (iii) designing a usability study to recognize the feasibility of VR-based social communication platform to map one’s physiological signals to his affective state of anxiety, thereby having implication on one’s performance.
Methods:
VR-based social conversation system was designed which projected humanoid characters with real-life faces having an Indian look, as classmates of the participant. These avatars moved dynamically in context-relevant VR environment while displaying gestures and narrating social stories in regional languages to the participants during task presentation. Also they demonstrated different context-relevant emotions e.g., happy, angry and neutral. Different difficulty levels depending on the type of questions asked by the avatars following the task presentation were also a part of the system design. Simultaneously the participants’ physiological signals e.g., heart rate, skin temperature, etc. were monitored via wireless wearable biosensors. A therapist rated the participants on their anxiety level. The participants’ performance was assessed based on their response to questions asked by the avatars.
Results:
Two children with ASD and three typically developing (TD) children participated in a usability study. Results indicate that both ASD and TD groups showed variations in their physiological signals. These variations had implication towards their performance. For ASD group the variations were high compared to their TD counterparts. This can be attributed to their high anxiety level as rated by the therapist during the study.
Conclusions:
The work presented here, shows the feasibility of a VR-based social conversation module that can switch tasks of varying difficulty based on one’s performance measure has the capability to predict participants’ anxiety level from their physiological markers. The preliminary data analysis shows the potential of developing a physiology based anxiety-sensitive intelligent adaptive VR platform to address the social communication challenges faced by these children.