Development and initial validation of FishFinder as a machine learning-Based serious video game for objective ADHD screening in children aged 5–12
摘要
Early identification of attention deficit hyperactivity disorder (ADHD), one of the most common childhood psychiatric disorders, is crucial for timely and effective intervention. Traditional screening methods typically rely on subjective evaluations from parents and teachers, which can introduce bias and are difficult to implement on a large scale. The emergence of digital serious games offers a promising alternative, providing an objective, engaging, and scalable approach to assessing ADHD symptoms. This study aimed to develop and validate a video game (FishFinder) for the screening of ADHD in Children using objective measurement of the core symptoms of this disorder. The FishFinder measures attention and impulsivity through in-game performance and evaluates the child’s hyperactivity using smartphone motion sensors. This game was tested on 26 children with ADHD (14 boys and 12 girls) and 26 healthy children (13 boys and 13 girls) aged 5 to 12 years. A support vector machine (SVM) was employed to detect children with ADHD. This system showed 91% accuracy, 91% sensitivity, and 94% specificity in the case of using a combination of movement features and in-game features. Moreover, the accuracy of FishFinder was 96% in the case of using only in-game features and 88% in the case of using just movement features. The FishFinder demonstrated a strong ability to identify ADHD in children in this study. Thus, this game may be used as an affordable, accessible, and enjoyable method for the objective screening of ADHD.