A Novel Model for the Identification and Classification of Thyroid Nodules Using Deep Neural Network

The accurate diagnosis of a thyroid nodule is essential due to an increase in the number of cases of malignant thyroid nodules. This research work presents a novel model for the identification and classification of thyroid nodules using deep neural network (DNN). The proposed model works in the following four phases (i) data collection (ii) pre-processing (iii) feature extraction and (iv) classification using DNN. In this work, eight features namely mean, variance, standard deviation, skewness, contrast, correlation, energy, and homogeneity are extracted using intensity and grey-level co-occurrence matrix (GLCM) methods. Experiments have been conducted on public and collected datasets with tenfold cross validation and 50–50% hold out method to validate the performance of the proposed model. It is inferred from the results that the proposed model performs better with an accuracy of 90.9%, sensitivity of 91.75% and specificity of 89.87% on public dataset and an accuracy of 92.85%, sensitivity of 93.68% and specificity of 91.78% on the collected dataset. It is evident that the proposed model is competitive to other state-of-the-art models for classifying thyroid nodules.