Handwriting-based ADHD Detection for Children Having ASD using Machine Learning Approaches

Attention deficit hyperactivity disorder (ADHD) for children is one of the behavioral disorders that affect brain’s ability to control attention, impulsivity, and its prevalence has increased over time. The cure ADHD still unknown only early detection can improve quality life with ADHD. At same time, often suffer from various comorbidities like autism spectrum (ASD), major depressive (MDD), etc. Various researchers developed computational tools detect depending on handwriting text. Handwriting text-based systems are a specific language causes problems non-native speakers language. Moreover, very few considered other such as ASD, MDD, etc., in their studies children. In this study, patterns or drawing assumed an aspect identify/detect who have ASD using machine learning (ML)-based approaches. We collected samples 29 Japanese (14 coexisting 15 healthy children) pen tablet. asked each child draw two patterns, namely zigzag lines periodic (PL) tablet repeated them three times. extracted 30 statistical features raw datasets these were analyzed sequential forward floating search (SFFS) selected best combinations subsets features. Finally, fed into seven ML-based algorithms detecting These classifiers trained leave-one-out cross-validation evaluated performances accuracy, recall, precision, f1-score, area under curve (AUC). experimental results illustrated highest performance scores (accuracy: 93.10%; recall: 90.48%; precision: 95.00%; f1-score: 92.68%; AUC: 0.930) achieved by RF-based classifier PL predict task. This study will be helpful provide evidence possibility classifying having based patterns.