Neural Network Based Identification of Skeletal and Dental Abnormalities Using McNamara Cephalometric Analysis

Abstract

cephalometric radiography to study relationships between bony and soft tissue landmarks and can be used to diagnose facial growth abnormalities prior to treatment, in the middle of treatment to evaluate progress or at the conclusion of treatment to ascertain that the goals of treatment have been met. This research work addresses McNamara analysis for classification of patients. In this research work, the backpropagation neural network (BPNN), and generalized regression neural network (GRNN) classifiers are used and studied for the diagnosis of Cephalometric analysis. In this study patient's data are collected from Raja Muthiah Dental College & Hospital (RMDC&H), Faculty of Dentistry, Annamalai University, Annamalai Nagar, Cuddalore District, Tamilnadu, India. A total of 304 (male 109, female 195) patients case records were collected for this study. All the collected clinical data are used for classification. For training and testing the proposed models, patients data were separated by four fold cross validation. Based on McNamara analysis, experimental results show that BPNN provided achieving the performance of 93.01% of good classification results when compared to the GRNN model. The BPNN approach is feasible and was found to be achieving a performance of 93.01% of the correct detection of patients. From this one can conclude that the BPNN model is capable of reproducing the target output values with minimal error.