This book contains the texture analysis on ultrasound medical images using quantitative information of the tissue characterization and structure for the detection of thyroid nodule. Gray level co-occurrence based texture features being extracted from thyroid images are use to classify into risk of malignancy using support vector machine and two layer feed forward neural network. It's providing the objective method to physician for diagnosis of pathological diseases such as thyroid gland ultrasound images for detection of nodules malignancy risk. The classification of benign thyroid nodule or malignant thyroid nodule will provide the second opinion to the physician about treatment of patient's and thus, this will help the physician to minimize the misdiagnosis rate of pathological diseases.