Qiu Xun, He Tao, Wu Xingda, Wang Peng, Wang Xin, Fu Qiuyue, Fang Xianglin, Li Shaoxin, Li Ying
College of Medical Technology, Guangdong Medical University, Dongguan, China.
Department of Biology, School of Basic Medical Sciences, Guangdong Medical University, Dongguan, China.
J Biophotonics. 2022 Dec;15(12):e202200158. doi: 10.1002/jbio.202200158. Epub 2022 Sep 11.
Cutaneous melanoma is a skin tumor with a high degree of malignancy and fatality rate, the incidence of which has increased in recent years. Therefore, a rapid and sensitive diagnostic technique of melanoma cells is urgently needed. In this paper, we present a new approach using fiber optical tweezers to manipulate melanoma cells to measure their Raman spectra. Then, combined with Principal Component Analysis and Support Vector Machines (PCA-SVM) classification model, to achieve the classification of common mutant, wild-type and drug-resistant melanoma cells. A total of 150 Raman spectra of 30 cells were collected from mutant, wild-type and drug-resistant melanoma cell lines, and the classification accuracy was 92%, 94%, 97.5%, respectively. These results suggest that the study of tumor cells based on fiber optical tweezers and Raman spectroscopy is a promising method for early and rapid identification and diagnosis of tumor cells.
皮肤黑色素瘤是一种具有高度恶性和致死率的皮肤肿瘤，近年来其发病率呈上升趋势。因此，迫切需要一种快速、灵敏的黑色素瘤细胞诊断技术。在本文中，我们提出了一种利用光纤镊子操纵黑色素瘤细胞以测量其拉曼光谱的新方法。然后，结合主成分分析和支持向量机（PCA-SVM）分类模型，实现对常见突变型、野生型和耐药型黑色素瘤细胞的分类。从突变型、野生型和耐药型黑色素瘤细胞系中总共收集了30个细胞的150条拉曼光谱，分类准确率分别为92%、94%、97.5%。这些结果表明，基于光纤镊子和拉曼光谱的肿瘤细胞研究是一种用于肿瘤细胞早期快速识别和诊断的有前景的方法。
