| Abstrakt: |
Objectives: While studies examining the effectiveness of virtual reality (VR) systems in autism spectrum disorder (ASD) intervention have seen significant growth, research on their application as tools to improve assessment and diagnosis remains limited. This systematic review explores the potential of VR systems in speeding-up and enhancing the assessment process for ASD. Methods: We conducted a systematic search of peer-reviewed research to identify studies that compared characteristics of autistic and neurotypical participants performing tasks in virtual environments. Pubmed and IEE Xplore databases were searched and screened using predetermined keywords and inclusion criteria related to ASD and virtual reality, resulting in the inclusion of 20 studies. Results: Studies reviewed revealed that VR technologies may serve as a booster of ASD “traits” that might otherwise go unnoticed when using traditional tools. Specifically, results indicated that ASD individuals exhibited distinct behavioral nuances compared to typically developing participants across four main domains: communication and social interaction skills, cognitive functioning and neurological pattern, sensory and physiological processing, and motor behavior and body movements. Also, recent studies analyzed here underscored the potential of integrating machine learning with VR technologies to enhance accuracy in identifying ASD based on motor behavior, eye gaze, and electrodermal activity. Conclusions: The integration of VR technologies can complement traditional tools in ASD diagnosis, providing more objective and reliable assessment within a controlled, ecological, and motivating virtual environment. In addition, the reviewed literature suggests machine learning models combined with VR technologies may support phenotypic diagnosis, offering a more refined classification of ASD subgroups within immersive virtual contexts. |
Full text from SpringerLink