Prognosis in Early-Stage Non-Small-Cell Lung Cancer Using Support Vector Machine
The current standard of treatment for patients with Early-Stage Non-Small-Cell Lung Cancer (NSCLC) is surgical resection. However, there are considerable differences among the survival time of those patients after resecting their tumors completely even if their have same clinical behaviors. So if we can predict the postoperative survival time of NSCLC scientifically and accurately, we will be able to select patients for individualized therapy, which is very significant to increase the curative effect of NSCLC. To predict the outcome of each individual patient, Support Vector Machine is used for- prognostic predication using tissue microarray data. Based on the observation that both nominal and numerical features are used, we proposes a novel hybrid kernel function that can utilize these two types of features. Compared with prognosis using gene-microarray data, our method is more readily adaptable to clinical practice because genemicroarrays techniques have the requirements for fresh or snap-frozen tissue and the examination costs is expensive. The simulation results are given to verify the efficiency and effectiveness of our proposed new method.
Prognostic Prediction Support Vector Machine Feature Selection
Bing-Yu Sun Zhi-Hua Zhu
Institute of Intelligent Machines Chinese Academy of Sciences Hefei, Anhui, P.R.Chian, 230031 the Lung Cancer Research Center Sun Yat-sen University Guangzhou, Guangdong Province
国际会议
重庆
英文
1487-1490
2011-08-20（万方平台首次上网日期，不代表论文的发表时间）