摘要
目的初步建立基于血清拉曼光谱的丙型肝炎病毒(HCV)的诊断及1b亚型的鉴定方法。方法应用激光拉曼光谱仪检测各样品的血清,检测波长区间为300至3 000 cm^(-1),采用airPLS-pls-svc算法处理拉曼光谱数据并分型鉴定,对降维后的主成分数据使用SVM(支持向量机)算法进行判别分析,建立训练集和测试集,其中HCV诊断训练集,纳入145例健康志愿者和152例HCV患者,HCV诊断测试集,纳入55例健康志愿者和72例HCV患者。1b亚型鉴定的训练集为88例1b亚型患者和64例非1b亚型患者,1b亚型鉴定的测试集为51例1b亚型患者和15例非1b亚型患者。结果 HCV患者和健康志愿者血清拉曼光谱拉曼峰型相似,在1 007、1 155、1 508 cm^(-1)波长的拉曼光谱峰强度存在差异,应用拉曼光谱诊断HCV的特异性和敏感性分别为98.611 1%(71/72)和70.909 1%(85/91),ROC曲线下面积(AUC)为0.981 7,1b型鉴定准确率为100%(51/51),非1b亚型鉴定准确率为93.333 3%(14/15),误诊一例。总准确率为98.484 8%,总误诊1例。结论本研究初步证明拉曼光谱可应用于HCV诊断和HCV1b亚型的鉴定,为研发HCV新型诊断和分型技术奠定了基础。
Objective To establish a method for diagnosis and identification of subtype 1 b hepatitis c virus(HCV) based on serum Raman spectra. Methods The serum of each group were detected by laser Raman spectrometer with the detection wavelength range of 300-3 000 cm-1. The Raman spectra data and classification identification was analyzed by the airPLS-PLS-SVC algorithm. The main component of data after the dimension reduction was processed by using the support vector machine(SVM) algorithm to discriminant analysis, and training set and test set was established. The training set for HCV diagnosis, included 145 cases of healthy volunteers and 152 patients with HCV. The test set for HCV diagnosis, included 55 cases of healthy volunteers and 72 patients with HCV. The training set for subtype 1 b identification was 88 cases of subtype 1 b patients and 64 cases of non-subtype 1 b patients, while the test set for subtype 1 b identification was 51 cases of subtype 1 b patients and 15 cases of non-subtype 1 b patients. Results The differential serum Raman peak between HCV patients and healthy volunteers type, were appeared in 1 007, 1 155, 1 508 cm-1 of the wavelength. The specificity and sensitivity of the Raman spectroscopy method for diagnosis the HCV were 98.611 1%(71/72) and 70.909 1%(85/91), and the area under the ROC curve for the AUC=0.981 7. The accuracy rate of Raman spectroscopy method for identification of 1 b type was 100%(51/51), and identification of non 1 b subtype was 93.333 3%(14/15). The total accuracy rate was 98.4848%, with a total misdiagnosis of 1 case. Conclusion This study preliminarily proved that Raman spectroscopy could be applied to the diagnosis of HCV and the identification of HCV subtype 1 b, which laid a foundation for the development of new diagnostic and typing techniques for HCV.
作者
潘琨琨
张朝霞
秦洁
吕小毅
PAN Kunkun;ZHANG Zhaoxia;QIN Jie;LV Xiaoyi(First Clinical Medical College, Xinjiang Medical University, Urumqi 830011, China;Department of Clinical Laboratory, the First Affiliated Hospital of Xinjiang Medical University, Urumqi 830054, China;Xinjiang University, Urumqi 830000, China)
出处
《新疆医科大学学报》
CAS
2019年第5期656-662,共7页
Journal of Xinjiang Medical University
基金
国家高技术研究发展计划(863计划)子课题(2015AA021107)
关键词
丙型肝炎病毒
基因型
拉曼光谱
hepatitis c virus
genotype
Raman spectroscopy
