Search

To date, the diagnostic tools for autism spectrum disorder (ASD) have been mostly based on qualitative criteria from observational information in contexts with low ecological validity. We are witnessing a growing scientific activity that proposes the use of implicit measures for the evaluation and diagnosis of ASD. These measures are based on processes of a biological and unconscious nature, underlying the capacity of human cognition, and are obtained through the acquisition and treatment of brain, physiological and behavioral responses in order to obtain the behavioral structure of the ASD patient facing a stimulus. The complex relationship between physiological responses and the behavioral structure of the ASD patient requires the use of advanced techniques of signal processing based on cognitive computation. Artificial intelligence (AI) techniques, such as machine learning and neurocomputing applied to the analysis of psychophysiological signals, have demonstrated their robustness for the classification of complex cognitive constructs. Virtual reality (VR) is a tool that allows recreating real-life situations with high sensory fidelity, but at the same time individually controlling each of the situations and stimuli that influence human behavior. It also allows the measurement in real time of human reactions to such stimuli. This document analyzes the latest scientific and technological advances relevant to its applications in the diagnosis of ASD. We conclude that VR is a very valuable tool for ASD research, especially for the evaluation and diagnosis of complex skills and competencies.