Abstract
In this paper, Raman spectra of human serum were measured using Raman spectroscopy, then the spectra was analyzed by multivariate statistical methods of principal component analysis (PCA). Then linear discriminant analysis (LDA) was utilized to differentiate the loading score of different diseases as the diagnosing algorithm. Artificial neural network (ANN) was used for cross-validation. The diagnosis sensitivity and specificity by PCA-LDA are 88% and 79%, while that of the PCA-ANN are 89% and 95%. It can be seen that modern analyzing method is a useful tool for the analysis of serum spectra for diagnosing diseases.
© 2011 OSA/SPIE
PDF ArticleMore Like This
Xiaozhou Li, Tianyue Yang, Ting Yu, and Siqi Li
808727 European Conference on Biomedical Optics (ECBO) 2011
Xiaozhou Li, Xiangxiang Tian, Tianyue Yang, Ting Yu, and Siqi Li
808728 European Conference on Biomedical Optics (ECBO) 2011
Xiaozhou Li, Tianyue Yang, Ting Yu, Ruomin Sun, and Siqi Li
808721 European Conference on Biomedical Optics (ECBO) 2011