Mestrado em Engenharia Elétrica

URI Permanente para esta coleção

http://repositorio.ufes.br/handle/10/17619

Nível: Mestrado Acadêmico
Ano de início:
Conceito atual na CAPES:
Ato normativo:
Periodicidade de seleção:
Área(s) de concentração:
Url do curso:

Navegar

Navegando Mestrado em Engenharia Elétrica por Assunto "ADHD"

Agora exibindo 1 - 1 de 1
  • Nenhuma Miniatura disponível
    Item
    Development of serious games for neurorehabilitation of children with attention-deficit/hyperactivity disorder through neurofeedback
    (Universidade Federal do Espírito Santo, 2020-04-14) Machado, Fabiana Santos Vieira; Frizera Neto, Anselmo; https://orcid.org/0000000206873967; http://lattes.cnpq.br/8928890008799265; https://orcid.org/0000-0003-0996-8651; http://lattes.cnpq.br/2705690076290294; Andrade, Kleber de Oliveira; http://lattes.cnpq.br/1498251399219988; Palacios, Ester Miyuki Nakamura; https://orcid.org/0000-0002-2196-2490; http://lattes.cnpq.br/1236573205554833
    Attention Deficit Hyperactivity Disorder (ADHD) is a neuropsychiatric disorder, which is treated with the administration of psychostimulants, and cognitive behavioral approaches. The use of the medication may have side effects; thus, neurofeedback can be used to teach the user how to regulate their own brainwaves to help correct its dysfunctional patterns. NFB has a repetitive aspect and can be long lasting. Therefore, “serious games” can be integrated into training to maintain interest throughout the process. This research presents the development of a neurofeedback system based on serious games for neurorehabilitation of children with ADHD. For the brain data acquisition, a 20 dry-electrode wireless EEG headset. The Neurofeedback Space serious game, developed in Unity platform, was developed to be used throughout the NFB sessions and to provide immersion and engagement. In the preliminary validation and algorithm’ selection, 5 volunteers were asked to watch a video with attention and non-attention tasks. Then, an SVM classifier was trained and tested with the collected data to choose between five brain regions and four algorithm versions with the best performance. The best results were achieved by (training accuracy of 98.12% and test accuracy of 93.88%) region 4 (F3, F7, C3 and T3) and version 2. Two experiments with 5 volunteers each, 5 sessions, and with an interval of 2 months between them were done to test the complete NFB system. The main difference between experiments was to fix game level 2 so that all metrics could be compared between volunteers. Three evaluation metrics were used to analyze the improvement in the performance: game’s metric, brain signal’s frequency bands power metric, and concentrated attention test metric. The analyzes show a divergence between the results obtained in both experiments. All volunteers improved in performance in the concentrated attention test. In Experiment 1, the Volunteers improved in the performance of the game metrics and reported feeling more concentrated throughout the week. In Experiment 2, the results of attention, score and sustained attention had large oscillations. Brain data results were inconclusive. Also, fixing level 2 showed no learning effect as perceived in experiment 1. Therefore, it is not possible to declare conclusive results for the developed NFB system, even with the improvement of some metrics presented. Factors that may have affected NFB training are number of sessions, duration of serious game, number of volunteers, game genre, and EEG dry electrodes headset.