Human Dental Age and Gender Assessment from Dental Radiographs Using Deep Convolutional Neural Network

Abstract

Human gender and age identification play a prominent role in forensics, bio-archaeology, and anthropology. Dental images provide prominent indications used for the treatment or diagnosis of disease and forensic investigation. Numerous dental age identification techniques come with specific boundaries, namely minimum reliability and accuracy. Gender identification approaches are not widely researched, whereas the effectiveness and accuracy of classification are not practical and very minimal. Drawbacks in the existing system are considered in the formulation of the proposed approach. Deep learning approaches can effectively rectify issues of drawbacks in other classifiers. The accuracy and performance of a classifier are enhanced with the deep convolutional neural network. The fuzzy C-Means Clustering approach is used for segmentation, and Ant Lion Optimization is used for optimal feature score selection. The selected features are classified using a deep convolutional neural network (DCNN). The performance of the proposed technique is investigated with existing classifiers, and DCNN outperforms other classifiers. The proposed technique achieves 91.7% and 91% accuracy for the identification of gender and age, respectively.