恶性肿瘤相关改变检测方法的研究  

Research on the detection of malignancy associated changes

作　　者：高智勇[1] 龚健雅[1] 秦前清[1] 林家瑞[2] 

机构地区：[1]武汉大学测绘遥感信息工程国家重点实验室,武汉430079 [2]华中科技大学生物医学工程研究所,武汉430074

出　　处：《北京生物医学工程》2005年第1期24-28,共5页Beijing Biomedical Engineering

摘　　要：恶性肿瘤相关改变 (malignanceassociatedchanges,MAC)可以作为癌症的指示集 ,通过检测MAC可以判断肿瘤患者的属性。由于MAC现象比较复杂 ,因此传统的检测方法效果不是很理想。针对MAC检测的特点 ,本文引入一种新的模式识别方法———支持向量机作为MAC检测方法。通过对不同病人的胸水样本细胞的恶性肿瘤相关改变的实验研究验证 ,支持向量机方法对MAC的检测有较高的识别率。Malignancy associated changes can be looked as the indicator of cancer and can be used to diagnosis the status of cancer. Due to the complexity of MAC, it is not perfect to be detected traditionally. According to the property of MAC, we presented a new method, originated from pattern recognition, support vector machine. In this paper, some experiments were implemented using support vector machine to discriminate the properties of the cancer of patients. The results show that high accuracy rate can be gained through use of support vector machine.

关 键 词：改变 恶性肿瘤 肿瘤患者 检测方法 胸水 病人 癌症 MAC 支持向量机 识别率 

分 类 号：R730[医药卫生—肿瘤] TP391[医药卫生—临床医学]