• Korean Journal of Head & Neck Oncology
• /
• v.13 no.2
• /
• pp.206-212
• /
• 1997

H$\"{u}$rthle cell neoplasm of the thyroid gland is an uncommon, but potentially malignant lesion. However, in many instances, the malignant potential of the H$\"{u}$rthle cell neoplasm is very difficult to judge histologically. For this reason, the biologic behavior of this tumor and its optimal treatment have come under considerable debate in recent years. In order to review the clinicopathologic features of the H$\"{u}$rthle cell neoplasm and to determine its optimal treatment modalities, we studied 26 patients with path logical proof of H$\"{u}$rthle cell tumor from January 1987 to September 1997. We also performed an immunohistochemical study using the monoclonal antibodies against antigen CD34 for the angiogenic activity of this tumor and evaluated the differences of microvessel density(MVD) between benign and malignant tumors. The age of the patients ranged from 1 to 71 years with a mean of 44.2 years. There were 6 males and 20 females(M : F= 1 : 3.3). The accuracies of fine needle aspiration biopsy and frozen section were very low; 6.3% and 34.8%, respectively. There were 20 benign tumors and 6 malignant tumors(23.1%). All the malignant tumors were microinvasive(intermediate) type which had minimal capsular invasion and most of them(5 cases) were diagnosed postoperatively. Any specific clinicopathologic differences were not seen between benign and intermediate groups. Most of the cases had conservative surgeries(15 ipsilateral lobectomy-isthmusectomy, 7 subtotal thyroidectomy) while total thyroidectomy was performed in 4 cases. Of the cases with malignant tumor, 2 had ipsilateral lobectomy-isthmusectomy, 3 had subtotal thyroidectomy and the remaining 1 had total thyroidectomy. Mean size of the tumors was 3.0 cm(0.1- 8.5 cm) in the greatest diameter and multiple tumors were seen in 6 cases(23.1 %). During the follow-up period, only one recurrence(3.8%) of benign tumor occurred but distant metastasis or cause-specific death was seen in the benign or intermediate groups. Mean MVDs of the benign(n=13) and intermediate(n=6) groups were $121.7{\pm}35.3$ and $114.3{\pm}31.7$, respectively and there was no statistical significance between them. In conclusion, because of the low accuracies of fine needle aspiration biopsy and frozen section for the H$\"{u}$rthle cell neoplasm, the extent of surgery could be individualized based on permanent pathologic examination; Conservative surgery would be adequate for patients with benign or intermediate H$\"{u}$rthle cell neoplasm and total or near-total thyroidectomy for those with definite malignancy.

• Journal of the Korea Society for Simulation
• /
• v.15 no.2
• /
• pp.31-39
• /
• 2006

The visibility graph is a well-known method for efficient path-finding with the minimum search space modelling the game world. The generalized visibility graph is constructed on the expanded obstacle boundaries to eliminate the "wall-hugging" problem which is a major disadvantage of using the visibility graph. The paths generated by the generalized visibility graph are guaranteed to be near optimal and natural-looking. In this paper we propose the method to apply the generalized visibility graph efficiently for game characters who moves among static obstacles between varying start and goal points. Even though the space is minimal once the generalized visibility graph is constructed, the construction itself is time-consuming in checking the intersection between every two links connecting nodes. The idea is that we build the map for static obstacles first and then incorporate start and goal nodes quickly. The incorporation of start and goal nodes is the part that must be executed repeatedly. Therefore we propose to use the rotational plane-sweep algorithm in the computational geometry for incorporating start and goal nodes efficiently. The simulation result shows that the execution time has been improved by 39%-68% according to running times in the game environment with multiple static obstacles.