© 2023, IRJET | Impact Factor value: 8.226 | ISO 9001:2008 Certified Journal | Page 710
NEUROASSIST ADHD ANALYZER: A SMART APPLICATION FOR
RECOGNIZING ATTENTION DEFICIT HYPERACTIVITY DISORDER
(ADHD) LEVELS IN CHILDREN
Gamage L.S.
Department of Information
Technology
Sri Lanka Institute of Information
Technology
Sri Lanka
it19382036@my.sliit.lk
Shehan R.H.A.
Department of Information
Technology
Sri Lanka Institute of Information
Technology
Sri Lanka
it19987576@my.sliit.lk
Bambarandage T. M. L.
Department of Information
Technology
Sri Lanka Institute of Information
Technology
Sri Lanka
it19176116@my.sliit.lk
Anjali W.M.S.
Department of Information
Technology
Sri Lanka Institute of Information
Technology
Sri Lanka
it19989242@my.sliit.lk
Thelijjagoda S.
Department of Information
Technology
Sri Lanka Institute of Information
Technology
Sri Lanka
samantha.t@sliit.lk
Krishara J.
Department of Information
Technology
Sri Lanka Institute of Information
Technology
Sri Lanka
jenny.k@sliit.lk
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Attention Deficit Hyperactivity Disorder (ADHD) is
a common neurodevelopmental disorder that affects children.
Identifying children with ADHD is a challenging task, as it
requires specialized training and expertise. A novel approach
to identifying children with ADHD using artificial neural
networks (ANNs) and convolutional neural networks (CNNs)
was proposed in this research paper. The web application
"NEUROASSIST ADHD Analyzer" was developed to assist in the
identification of children with attention deficit hyperactivity
disorder (ADHD) using electroencephalogram (EEG) and
facial expression data. The application includes three focus
games, focus activities, a facial expression analyzer, a
combined probability predictor, and a BOT application for the
SNAP IV questionnaire. The ANN and CNN models were used to
implement these components. An LR model was used to predict
the overall result. The integration of these multiple models
into the combined model resulted in an accuracy of 94%, with
a sensitivity of 90% and a specificity of 92%. Additionally, the
area under the receiver operating characteristic curve (AUC-
ROC) was 0.91, indicating good discriminative power.
Furthermore, a user study indicated that the application is
user-friendly and effective in assisting with the diagnosis of
ADHD. This research demonstrates the potential of ANNs and
CNNs in identifying children with ADHD and presents a
promising tool for assisting clinicians in diagnosing the
disorder. Future work could focus on improving the accuracy
of the models by incorporating additional data sources and
features and expanding the application to a larger population.
Key Words: ADHD, EEG, CNN, ANN, machine learning
1. INTRODUCTION
ADHD or attention deficit hyperactivity disorder is a
condition that affects both children and adults and is
categorized as a neurodevelopmental disorder. The disorder
is characterized by a continuous pattern of inattention,
hyperactivity, and impulsivity that can significantly hinder
daily functioning and development. Inattention symptoms
include difficulties sustaining attention, making careless
mistakes, organizing tasks and activities, and following
through on instructions. Fidgeting, excessive chatting,
interrupting others, and difficulty waiting for one's turn are
only a few characteristics of hyperactivity-impulsivity [1].
ADHD is an important topic to research due to its high
prevalence and the detrimental effects it may have on people
as well as society at large. The National Institute of Mental
Health (NIMH) estimates that 9.4% of American children
between the ages of 2 and 17 have ADHD. The disorder can
persist into adulthood, with approximately 4.4% of US adults
estimated to have ADHD [2].
ADHD can lead to significant impairments in various areas of
functioning, including academic and occupational
achievement, social skills, and mental health. Individuals
with ADHD are at an increased risk of [3] experiencing
problems such as academic underachievement, relationship
difficulties, and substance abuse. Moreover, ADHD can cause
a significant financial burden, as individuals with the
disorder often require more medical care and may have
decreased earning potential [4].
International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 10 Issue: 06 | Jun 2023 www.irjet.net p-ISSN: 2395-0072