Detection and Classification of Multi-Type Cells by Using Confocal Raman Spectroscopy

Tumor cells circulating in the peripheral blood are the main cause of cancer metastasis and death, thus the identification and discrimination of these rare cells is valuable in the diagnosis of cancer. As a label-free detection method without invasion, Raman microscopy has already been proved as a promising method for cell identification. In this paper, we use a confocal Raman microscopy with 532nm laser excitation to obtain the Raman spectrum of six different kind of living cells such as 786-O, HKC, HepG-2, A549, A375, 4T1. Multivariate statistical methods are applied to classify the Raman spectra of these cells. The results show that these cells have been able to be distinguished from each other. The cell types can be predicted correctly by using quadratic discriminant analysis model with the highest accuracy of 100%. It is demonstrated that this analysis model based on Raman spectrum of cells is promising and will have a great potential in the biomedical fields of classifying circulating tumor cells in the future.