The prevalence of thyroid nodule is worldwide high. About 40% of normal population has thyroid nodules and about 5% are malignancy. It is important to differentiate malignancy from benign nodules because the management is completely different. Nowadays, the gold standard is fine needle aspiration cytology (FNAC) examination. The overall sensitivity and specificity is fair (\ 90%), but still has its limitation that some results are indeterminate in about 15% of the nodules. These obstacles are especially troublesome for papillary and follicular thyroid cancer that leads to delayed diagnosis, incomplete resection, and repeated operation. Preoperative evaluation of the prognosis is extraordinary important for cancer management. However, current prognostic scoring systems is only applicable after surgery. Hence, we urgently need a better risk-stratification system for individual-tailored treatment, and genetic-based computerized morphometry study seems to be the most realistic and promising one. The goal of this study is to propose a reliable method for diagnosis and prognosis of papillary thyroid cancer and follicular thyroid cancer through analyzing cellular morphologic characteristics.