论文部分内容阅读
目的应用拉曼光谱及支持向量机技术鉴别诊断口腔黏膜鳞状细胞癌,为拉曼光谱技术的临床应用提供理论基础。方法收集手术切除口腔黏膜鳞状细胞癌组织56例、上皮重度异常增生组织50例及正常口腔黏膜组织32例,应用拉曼光谱仪进行检测,获取不同组织的特征拉曼光谱。比较不同组织间拉曼光谱谱峰差异,并应用支持向量机技术对光谱数据进行分析,建立鉴别诊断模型。结果口腔黏膜不同组织间拉曼光谱差异主要集中在640、713、957、1002、1 077、1 306、1445和1 653 cm~(-1)等处,主要反映组织中蛋白质、核酸及脂类等生物大分子变化。支持向量机建立鉴别诊断模型的诊断准确率达94.3%~100.0%。结论口腔黏膜鳞状细胞癌组织、上皮重度异常增生组织及正常口腔黏膜组织间拉曼光谱存在差异,应用支持向量机建立诊断模型可以很好地鉴别诊断不同组织的光谱数据。
Objective To evaluate the diagnostic value of Raman spectroscopy and support vector machine in the diagnosis of oral mucosal squamous cell carcinoma (OSCC), and to provide a theoretical basis for the clinical application of Raman spectroscopy. Methods 56 cases of oral mucosal squamous cell carcinoma, 50 cases of epithelial dysplasia and 32 cases of normal oral mucosa were resected. Raman spectroscopy was used to detect the characteristic Raman spectra of different tissues. The differences of Raman spectrum peaks between different tissues were compared. The support vector machine (SVM) was used to analyze the spectral data to establish the differential diagnosis model. Results The differences in Raman spectra of different tissues of oral mucosa mainly focused on the areas of 640, 713, 957, 1002, 1 077, 1 306, 1445 and 1 653 cm -1, mainly reflecting the changes of protein, nucleic acid and lipid Other biological macromolecules change. The diagnostic accuracy rate of support vector machine in establishing differential diagnosis model is 94.3% ~ 100.0%. Conclusion The Raman spectra of oral mucosal squamous cell carcinoma, epithelial dysplasia and normal oral mucosa are different. Using the support vector machine to establish the diagnostic model can well differentiate the different tissues’ spectral data.