An enhanced anarchic society optimization technique for the classification of ultrasound thyroid images using ILBP

In the recent times, Thyroid Nodules (TNs) is a generic cancer of the thyroid gland, which impacts close to 20% of the population worldwide and nearly 50% of 60- year-old individuals. The conventional diagnostic method, relying on the expertise of doctors, shows a huge drawback that the diagnosis result very much relies on the individual knowledge and experience of the physician. As a result, efficacy of diagnosis is confined, and it varies with the doctor’s experience.  To combat this limitation, an efficient double screening technique is employed in few health care centers and hospitals by using one more specialist but, this approach is unaffordable and its time complexity is high. The research classified the thyroid nodules employing different image preprocessing techniques.  Utilized histogram equalization for preprocessing in his work. The Gray-Level Co-Occurrence Matrix (GLCM) is deployed for extracting the significant features. The classification is done using ASO, k-Nearest Neighbor (KNN), and Bayesian. It is noticed that the ASO yields improved accuracy compared to KNN and Bayesian techniques