PANDAS: Paediatric Attention-Deficit/Hyperactivity Disorder Application Software

Attention-deficit/hyperactivity disorder (ADHD) is a common neuropsychiatric disorder that impairs social, academic, and occupational functioning in children, adolescents and adults. In South Africa, youth prevalence of ADHD is estimated as 10%. It is therefore necessary to further investigate metho...

Full description

Saved in:

Bibliographic Details
Published in:	2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) Vol. 2019; pp. 1444 - 1447
Main Authors:	Mwamba, Herve M., Fourie, Pieter R., den Heever, Dawie van
Format:	Conference Proceeding Journal Article
Language:	English
Published:	United States IEEE 01-07-2019
Subjects:	Adolescent Attention Deficit Disorder with Hyperactivity Child Child, Preschool Feature extraction Games Humans Pediatrics Presses Prevalence Sample Size Sensitivity Software Support Vector Machine Support vector machines
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Description
Summary:	Attention-deficit/hyperactivity disorder (ADHD) is a common neuropsychiatric disorder that impairs social, academic, and occupational functioning in children, adolescents and adults. In South Africa, youth prevalence of ADHD is estimated as 10%. It is therefore necessary to further investigate methods that objectively diagnose, treat, and manage the disorder. The aim of the study was to develop a novel method that could be used as an aid to provide screening for ADHD. The study comprised of a beta-testing phase that included 30 children (19 non-ADHD and 11 ADHD) between the ages of 5 and 16 years old. The strategy was to use a tablet-based game that gathered real-time user data during game-play. This data was then used to train a linear binary support vector machine (SVM). The objective of the SVM was to differentiate between an ADHD individual versus a non-ADHD individual. A feature set was extracted from the gathered data and sequential forward selection (SFS) was performed to select the most significant features. The test set accuracy of 85.7% and leave-one-out cross-validation (LOOCV) accuracy of 83.5% were achieved. Overall, the classification accuracy of the trained SVM was 86.5%. Finally, the sensitivity of the model was 75% and this was seen as a moderate result. Since the sample size was fairly small, the results of the classifier were only seen as suggestive rather than conclusive. Therefore, the performance of the classifier was indicative that a quantitative tool could indeed be developed to perform screening for ADHD.
ISSN:	1557-170X 1558-4615
DOI:	10.1109/EMBC.2019.8857357