Age estimation based on length of left hand bone in African American children below 18 years old using Artificial Neural Network

• Main Author: Nor Anis Nabila, Saari
• Format: Undergraduates Project Papers
• Language: English
• Published: 2019
• Subjects: QA75 Electronic computers. Computer science; QA76 Computer software
• Online Access: http://umpir.ump.edu.my/id/eprint/27084/1/18.Age%20estimation%20based%20on%20length%20of%20left%20hand%20bone%20in%20African%20American%20children%20below%2018%20years%20old%20using%20artificial%20neural%20network.pdf; http://umpir.ump.edu.my/id/eprint/27084/; http://fypro.ump.edu.my/ethesis/index.php
• Institution: Universiti Malaysia Pahang

Age estimation is used in the field of forensic anthropology’s studies to assists in the identification of individual’s identity. The age estimation using traditional method was unique and applicable for a certain population only. The focus on this study is the measurement of left hand bone to estimate age using mathematical method of Multiple Linear Regression and also the soft computing models of Artificial Neural Network (ANN) that can contribute to another alternative models instead of using the traditional model of Greulich and Pyle (GP) model and Tanner and Whitehouse (TW) model that is based on the expert of anthropology’ s experience which may lead to various of result of age estimation. The regression models were carried out on X-ray images of the left hand in African American dataset from new-born to 18 years old. All the nineteen bones of the left hand were measured manually using Photo Pos, Power of Software Company Ltd which is the free photo editor that will creating a line on the each of left-hand bones. For Artificial Neural Network to produce a better result in prediction of age, hidden neuron network in ANN is manipulated as suggested by Zain et al. using Encog Workbench tools version 3.3.0. The value of R-square and mean square error (MSE) of proposed model been calculated as performance measurement for compare with benchmark of age. Based on result produced by these models, mean square error produced by ANN model are 1.775 and 2.487 for both male and female, respectively. To conclude, ANN is reliable to estimate the age based on length of the left hand.