Multivariate statistical image processing for molecular specific imaging in organic and bio-systems

Processing TOF-SIMS images to obtain clear contrast between chemically distinct regions, distinguish between chemical and topographic effects and identify chemical species can be a formidable challenge, particularly when working with organic and biological molecules that have similar spectral features. Three multivariate statistical techniques, including principal components analysis (PCA), multivariate curve resolution (MCR), and maximum auto-correlation factors (MAF) have been explored to determine their utility for processing TOF-SIMS images. The methods have been exhaustively tested on synthetic images to allow quantitative assessment of their utility. The methods are compared here based on enhancement of image contrast, enhancement of image resolution, and isolation of pure component spectra. MAF, which includes information on the nearest neighbors to each pixel, shows clear advantages over PCA and MCR for enhancing image contrast and identifying sparse components in the matrix. However, MCR is better suited to identification of unknown compounds. No single method proves superior for all of these objectives so a simple strategy is presented for combining these methods to obtain optimal results.