Detalhes bibliográficos
| Ano de defesa: |
2021 |
| Autor(a) principal: |
ARAÚJO, Claudyane da Silva |
| Orientador(a): |
CARVALHO, Ewaldo Eder Santana |
| Banca de defesa: |
SANTANA, Ewaldo Eder Carvalho,
SOUSA, Nilviane Pires Silva,
BARROS FILHO, Allan Kardec Duailibe,
ROSA, Claudia R. de Andrade A. |
| Tipo de documento: |
Dissertação
|
| Tipo de acesso: |
Acesso aberto |
| Idioma: |
por |
| Instituição de defesa: |
Universidade Federal do Maranhão
|
| Programa de Pós-Graduação: |
PROGRAMA DE PÓS-GRADUAÇÃO EM ENGENHARIA DE ELETRICIDADE/CCET
|
| Departamento: |
DEPARTAMENTO DE ENGENHARIA DA ELETRICIDADE/CCET
|
| País: |
Brasil
|
| Palavras-chave em Português: |
|
| Palavras-chave em Inglês: |
|
| Área do conhecimento CNPq: |
|
| Link de acesso: |
https://tedebc.ufma.br/jspui/handle/tede/3410
|
Resumo: |
The prevalence of overweight in adolescence is a worldwide public health problem, as it is associated with several metabolic disorders, such as cardiovascular diseases and diabetes. Such problems if not evaluated and treated early can lead to negative outcomes such as premature death, so the importance of analyzing the body fat of this population. Thus, the objective of this study is to develop an artificial neural network (ANN) to predict the percentage of body fat (%GC) of adolescents. In this network are used, as input parameters, weight, height, age, gender, heart rate, waist circumference, hip circumference, arm circumference. For training and testing of ANN, we used 5-fold cross-validation in a set of data from 772 adolescents of both sexes, aged between 10 and 19 years. For data labeling, we used the (%GC) obtained by bioimpedance (BIA). The prediction given by our RNA was compared with the prediction of other anthropometric methods commonly used in the evaluation of nutritional status. When comparing the value obtained by the net, in the test phase, with the value of BIA a correlation R= 0.87 was obtained. Our method showed significantly better results than the usual anthropometric indicators such as Body Mass Index (BMI), Waist-Height Relationship (WHtR), as can be evaluated by the area over the ROC curve (AUROC):0.83 (RNA), 0.62 (BMI) and 0.56 (WHtR). Our proposal also obtained 85.3% accuracy, 73.2% specificity, the sensitivity of 93%, and a 59.09% rate of true positives. These results are much better than the BMI and CER methods that present low sensitivity (27.6% and 11.2%). The specificity of our method showed a high rate of true negatives (26.28%). Thus, it is concluded that the RNA model obtained a better performance to predict excess body fat in adolescents compared to the usual anthropometric indicators, presenting itself as a low cost alternative for the tracking of obesity in adolescents. |
