Digital Dewaxing of Raman Signals: Discrimination Between Nevi and Melanoma Spectra Obtained from Paraffin-Embedded Skin Biopsies

摘要: Malignant melanoma (MM) is the most severe tumor affecting skin and accounts for three quarters of all cancer deaths. Raman spectroscopy a promising nondestructive tool that has been increasingly used characterization molecular features cancerous tissues. Different multivariate statistical analysis techniques are in order to extract relevant information can be considered as functional spectroscopic descriptors particular pathology. Paraffin embedding (waxing) highly efficient process conserve biopsies banks several years. However, use non-dewaxed formalin-fixed paraffin-embedded tissues investigations remains very restricted, limiting development technique routine analytical biomedical purposes. This due intense signal paraffin, which masks important vibrations biological In addition being time consuming chemical intensive, dewaxing methods not they leave traces paraffin tissues, affects signal. present study, we independent component (ICA) on spectral images collected nevus samples. The sources obtained from these then eliminate, using non-negativity constrained least squares (NCLS), contribution each individual spectrum nevi melanomas. Corrected spectra both types lesion compared classified into dendrograms hierarchical cluster (HCA).