Fuzzy Clustering of Raman Spectral Imaging Data with a Wavelet-Based Noise-Reduction Approach

Abstract

Raman spectral imaging has been widely used for extracting chemical information from biological specimens. One of the challenges is to cluster the chemical groups from the vast amount of hyperdimensional spectral imaging data so that functionally similar groups can be identified. In this paper, we present an approach that combines a differential wavelet-based data smoothing with a fuzzy clustering algorithm for the classification of Raman spectral images. The preprocessing of the spectral data is facilitated by decomposing them in the differential wavelet domain, where the discrimination of true spectral features and noise can be easily performed using a multi-scale pointwise product (MPP) criterion. This approach is applied to the classification of spectral data collected from adhesive/dentin interface specimens where the spectral data exhibit different signal-to-noise ratios. The proposed wavelet approach has been compared to several conventional noise-removal algorithms.