Home // International Journal On Advances in Life Sciences, volume 13, numbers 1 and 2, 2021 // View article
Authors:
Anderson Silva
Luiz Carreiro
Mayara Silva
Maria Teixeira
Leandro Silva
Keywords: Self-Organizing Maps (SOM); Decision Tree; Attention Deficit/Hyperactivity Disorder (ADHD)
Abstract:
Attention-Deficit/Hyperactivity Disorder (ADHD) presents in children and adolescents as a persistent pattern of inattention, hyperactivity, and impulsivity that interferes with their development. Computational studies on ADHD focus on measures of brain activity of the participants and a few use standardized cognitive tests or behavioral inventories to assess objective indicators for diagnosis. The paper presents a computational proposal in which the combination of two artificial intelligence methods is used to aid the identification of diagnostic indicators for ADHD. The proposal is to combine a neural network of self-organizing maps to group factors from standardized tests and inventories, and a decision tree to classify the most relevant factors. The study included 127 children and adolescents from 6 to 16 years old, 48 with ADHD diagnosis and 79 without ADHD (control group). The most relevant result of the study was the strong contribution of the Child and Adolescent Behavior Inventory results in the diagnosis of the disorder with great performance in prediction when compared to real data and reliability by Kappa statistics.
Pages: 54 to 64
Copyright: Copyright (c) to authors, 2021. Used with permission.
Publication date: December 31, 2021
Published in: journal
ISSN: 1942-2660