FTIR as a new approach to screen a rare disease.

This study investigated Raman spectroscopy and supervised classification for CLL diagnosis.

The hierarchical double application of crisp cluster validity indices for automated spectral histology of a normal human colon.

Raman spectra are classically modeled as a linear mixing of spectra of molecular constituents of the analyzed sample. Source separation methods are thus well suited to estimate these constituent spectra. However, physical distortions due to the instrumentation and biological nature of samples add nonlinearities to the Raman spectra model. These distortions are dark current, detector and optic responses, fluorescence background, and peak misalignment and peak width heterogeneity. The source separation results are thus deteriorated by these effects. We propose to develop specific preprocessing steps to correct these distortions and to retrieve a linear model. The benefits brought by these steps are studied by the application of two different source separation methods named joint approximate diagonalization of eigenmatrices and maximum likelihood positive source separation after the application of each step on a dataset acquired on a paraffin-embedded human skin biopsy. The efficacy of these methods to separate Raman spectra is also discussed.

Raman spectroscopy is a candidate technique for diagnosis applications in medicine due to its high molecular specificity. Optimizing the pre-treatment applied for Raman data is important for exploiting Raman signals and ensuring their relevance in medical diagnosis. One of the crucial steps in data pre-processing, normalization, can affect significantly the result interpretation. To select the appropriate normalization method, a strategy based on validity indices (VI) is proposed in this study. VI are based on measuring the quality of data partitioning without involving a full sequence of supervised classification. The approach was tested on Raman data acquired from control and in vitro glycated proteins (albumin and collagen). Protein glycation is a process involved in the molecular ageing of tissues that leads to the formation of products altering the functional and structural properties of proteins. Different methods of normalization were applied on the data sets: integrated intensity of the phenylalanine band, integrated intensity of the amide I band, standard normal variate (SNV), multiplicative signal correction (MSC), and extended multiplicative signal correction (EMSC) that performs simultaneously baseline correction and normalization. Following normalization, principal component analysis (PCA) was applied and VI were calculated from PCA scores resulting from each of the normalization methods mentioned. Based on VI quantitative values, our experiments permit to illustrate the effect of normalization on the data separability of control and glycated samples, and to determine the most appropriate normalization and simultaneously the most discriminant principal components to exploit vibrational information associated with glycation-induced modifications. In parallel, principal component analysis – linear discriminant analysis (PCA-LDA) was carried out for positioning the interest of VI in regard to a common chain of data processing.