Differential diagnosis of thyroid nodules with ultrasound elastographybased on support vector machines

A fine needle aspiration (FNA) biopsy is the standard procedure of choicefor differentiating between benign and malignant thyroid nodules, and 300,000thyroid FNA biopsies are performed in the U.S. each year. However, FNA isinvasive, costly and uncomfortable for patients. Furthermore, a large percentage(70%) of these FNAs turn out to be benign. In this paper, we present anon-invasive and automatic approach for differentiating benign and malignantthyroid nodules with ultrasound elastography based on support vector machines(SVM) with biased penalties. We used the elastography data of 98 thyroid nodules(16 malignant and 82 benign) from 92 subjects previously acquired with aclinical ultrasound machine, Hitachi Hi Vision 5500. We conducted theleave-one-out cross-validation (LOOCV) in evaluating the performance of ourclassification method. Our goal was to obtain the maximum geometric mean (MGM)of sensitivity and specificity. The results show that our method was able to getMGM of 90.1% with the sensitivity of 93.8% and the specificity of 86.6%.