Esteves B, Pimenta S, Maciel M J, Costa M, Baltazar F, Cerqueira M F, Alpuim P, Silva C A, Correia J H
CMEMS-UMinho, Department of Industrial Electronics, University of Minho, Guimarães, Portugal.
LABBELS - Associate Laboratory, Braga, Guimarães, Portugal.
Heliyon. 2024 Aug 28;10(17):e36981. doi: 10.1016/j.heliyon.2024.e36981. eCollection 2024 Sep 15.
This paper demonstrates the potential of Raman spectroscopy for differentiating neoplastic from non-neoplastic colon tumors, obtained with the CAM (chicken chorioallantoic membrane) model. For the CAM model two human cell lines were used to generate two types of tumors, the RKO cell line for neoplastic colon tumors and the NCM460 cell line for non-neoplastic colon tumors. The Raman spectra were acquired with a 785 nm excitation laser. The measured Raman spectra from the CAM samples ( = 14) were processed with several methods for baseline correction and to remove artifacts. The corrected spectra were analyzed with PCA (principal component analysis). Additionally, machine learning based algorithms were used to create a model capable of classifying neoplastic and non-neoplastic tumors. The principal component scores showed a clear differentiation between neoplastic and non-neoplastic colon tumors. The classification model had an accuracy of 93 %. Thus, a complete methodology to process and analyze Raman spectra was validated, using a rapid, accessible, and well-established tumor model that mimics the human tumor pathology with minor ethical concerns.
本文展示了拉曼光谱法在鉴别肿瘤性与非肿瘤性结肠肿瘤方面的潜力，该研究是通过鸡胚绒毛尿囊膜（CAM）模型获得的。对于CAM模型，使用了两种人类细胞系来生成两种类型的肿瘤，用于肿瘤性结肠肿瘤的RKO细胞系和用于非肿瘤性结肠肿瘤的NCM460细胞系。拉曼光谱是用785nm激发激光采集的。对来自CAM样本（n = 14）的测量拉曼光谱采用了几种方法进行基线校正和去除伪影。对校正后的光谱进行主成分分析（PCA）。此外，基于机器学习的算法被用于创建一个能够区分肿瘤性和非肿瘤性肿瘤的模型。主成分得分显示肿瘤性和非肿瘤性结肠肿瘤之间有明显差异。分类模型的准确率为93%。因此，使用一个快速、可及且成熟的肿瘤模型，在伦理问题较少的情况下模拟人类肿瘤病理学，验证了一种处理和分析拉曼光谱的完整方法。
