American and British scientists conducted a cross-sectional study and concluded that assessing movement imitation using an experimental video game improves the accuracy of differential diagnosis of autism spectrum disorders (ASD). The report was published in The British Journal of Psychiatry.
Currently, there are no reliable and specific biomarkers in clinical practice to distinguish autism spectrum disorders (ASD) from another common and often associated condition, attention deficit hyperactivity disorder (ADHD). This complicates diagnosis and timely access to treatment, which negatively affects the quality of life of patients.
Romila Santra from the Kennedy-Krieger Institute and colleagues from the UK and the US recruited 183 children aged 7–13 years to participate in the study. Of these, 35 had ADHD without ASD, 63 had ASD with concomitant ADHD, 21 had ASD without ADHD, and the remaining 65 were neurotypical (without neuropsychiatric features). The diagnoses corresponded to the DSM-5 criteria. All participants had preserved intelligence (IQ 70 or more). The children's condition was assessed using the standardized ADOS-2 test, the SRS-2, Conners-3, and Conners-4 parent questionnaires, and the PANESS behavioral assessment with and without CAMI. CAMI (Computerized Assessment of Motor Imitation) is an experimental method that involves asking a child to imitate dance-like movements with their entire body following an avatar on the screen during two minute sessions. During this time, the child is filmed in two projections with video cameras and the average accuracy of imitation is calculated from 0 to 1.
It turned out that the accuracy of imitation in CAMI was comparable in neurotypical and ADHD children, while it was significantly reduced in children with ASD, regardless of the presence of concomitant ADHD (p < 0.0001; R2 = 0.28). Analysis using the ROC curve and the support vector machine showed that the technique significantly helps to distinguish children with ASD from both neurotypical (the proportion of true positive cases was 80 percent) and children with ADHD (the proportion of true positive cases was 70 percent). Among participants with ASD, low CAMI scores were associated with more pronounced autistic features, in particular the level of social affect, restricted and repetitive behavior patterns on the ADOS-2, but not with ADHD features or motor skills. Thus, computerized assessment of motor imitation can serve as a relatively simple and objective additional method of differential diagnosis of ASD, including when differentiating them from ADHD, the authors of the study conclude. Previously, it was suggested to use gaze tracking with a smartphone app, blood analysis for the adipokine FABP4, and processing of functional connectivity maps of the brain using machine learning algorithms in diagnosing ASD. More detailed information about ASD can be found in the material "Children of the Rain".
