ABSTRACT
Attention deficit/hyperactivity disorder (ADHD) is a childhood-onset neurodevelopmental disorder in primary care settings. ADHD is categorized into three subtypes, namely inattention also known as ADHD-I, hyperactivity impulsivity, or ADHD-H, and combined, ADHD-C (a fusion with features of ADHD-H and ADHD-I), which are developmentally inconsistent with the age of the child. The worldwide research revolving around ADHD can be divided into different scopes of studies or fields, such as cause or etiology, prevalence, and case studies, policies and reviews, assessments and rating scales, treatment or aid, and diagnosis or detection. There are two types of sensors used in the detection of the ADHD-HI subtype, namely wearable and nonwearable. The wearable sensors are actigraph, IMUs, accelerometers, and pedometers, while the nonwearable sensors are infrared motion analysis and radio ultrawideband radar sensors. The analysis is recorded as the split signal in the windows with the same size using convolutional neural networks (CNN) architecture to classify the spectrograms of activity windows.
