基于拉曼光谱的卵巢癌诊断研究

发布时间：2018-06-04 02:18

本文选题：卵巢癌 + 拉曼光谱　；参考：《光谱学与光谱分析》2017年06期

【摘要】：卵巢癌是一种发病率和致死率极高的女性妇科疾病。目前卵巢癌的临床诊断主要依靠病理学检测,超声法以及检测血液中肿瘤标志物CA125,但是上述几种方法都存在其固有的缺陷。本研究提出应用拉曼光谱结合偏最小二乘-判别分析(PLS-DA)模型,实现在分子水平上判别诊断卵巢癌。拉曼光谱在正常组织与癌组织之间存在反映其癌变过程物质结构变化的微小差异。因此,通过结合PLS-DA数据模型分析拉曼光谱信息,能够将微小差异放大化,捕获生物分子的显著特征。在研究中,模型的变量数目选择其分类错误率最小时的数值,即隐变量的数目为5,能够捕获大量的官能团特征性信息,通过分析隐变量的p值大小可知5个隐变量均可实现对正常组织与癌组织的有效区分,并且第一个隐变量具有最明显的区分结果。通过模型运算结果可知,该模型对卵巢癌判别的准确性达到85.2%(其中灵敏性为86.2%,特异性为85.4%)。研究结果表明,拉曼光谱技术,通过与PLS-DA模型相结合,可作为卵巢癌临床诊断中的辅助诊断方法,从分子水平上实现对卵巢癌的诊断判别。
[Abstract]:Ovarian cancer is a female gynecological disease with high morbidity and mortality. At present, the clinical diagnosis of ovarian cancer mainly depends on pathological examination, ultrasound method and detection of tumor marker CA125 in blood, but these methods all have their inherent defects. In this study, Raman spectroscopy combined with partial least squares discriminant analysis (PLS-DAA) model was proposed to discriminate and diagnose ovarian cancer at molecular level. There is a slight difference in Raman spectra between normal and cancerous tissues reflecting the changes of substance structure during carcinogenesis. Therefore, by analyzing the Raman spectrum information with PLS-DA data model, we can amplify the small differences and capture the remarkable characteristics of biomolecules. In the study, the number of variables in the model is chosen to be the value of the minimum classification error rate, that is, the number of hidden variables is 5, which can capture a large amount of functional group characteristic information. By analyzing the p value of the hidden variable, we can know that the five hidden variables can effectively distinguish the normal tissue from the cancer tissue, and the first hidden variable has the most obvious result. The results show that the accuracy of the model is 85.2 (the sensitivity is 86.2 and the specificity is 85.4). The results show that Raman spectroscopy, combined with PLS-DA model, can be used as an auxiliary diagnostic method in the diagnosis of ovarian cancer, and it can be used to diagnose ovarian cancer at molecular level.
【作者单位】：环境化学与生态毒理学国家重点实验室中国科学院生态环境研究中心中国科学院大学;山东大学环境科学与工程学院;
【基金】：中国科学院战略性先导专项(XDB14020201)资助
【分类号】：O657.37;R737.31

【相似文献】