Nutritional assessment is an important evaluation to prevent and control malnutrition, which is one of the main causes associated with child mortality. Weight and height are the most frequently measured morphological traits which in combination with child's gender and age, generates anthropometric indices to establish child's nutritional status. Nevertheless, accomplishment of this task in rural areas is difficult because of complications to transport bulky and heavy equipment, which must be properly and adequately calibrated. This work proposed a novel approach to perform nutritional assessments of children under five, through a system focused on the estimation of anthropometric indices, based on the measurements obtained from a set of body part images and its relations with child's gender and age. The results showed that sensitivity and specificity for the anthropometric indicators, ranged from 66% to 100% and 88% to 100%, respectively. Moreover, overall accuracies were over 85% up to 100%. Additionally, the experiments conducted shown our method as a viable solution to perform nutritional evaluations via accurate anthropometric index estimations.
|Title of host publication||Proceedings of the 2016 IEEE ANDESCON|
|Place of Publication||United States|
|State||Published - 27 Jan 2017|
|Event||SIBGRAPI Conference on Graphics, Patterns and Images - Sao Paulo, Brazil|
Duration: 4 Oct 2016 → 7 Oct 2016
Conference number: 29
|Name||Brazilian Symposium on Computer Graphics and Image Processing|
|Conference||SIBGRAPI Conference on Graphics, Patterns and Images|
|Period||4/10/16 → 7/10/16|
Bibliographical notePublisher Copyright: © 2016 IEEE.
SIBGRAPI - Conference on Graphics, Patterns and Images is an international conference annually promoted by the Brazilian Computer Society (SBC).
Date Added to IEEE Xplore: 02 February 2017.
- image processing
- machine learning
- neural networks
- nutrition assessment