Imaging mass spectrometry based exploration of biochemical tissue composition using peak intensity weighted PCA

Imaging mass spectrometry or mass spectral imaging (MSI) is a technology that provides us with the opportunity to study the spatial distribution of biomolecules such as proteins, peptides, and metabolites throughout organic tissue sections. MSI adds a spatial dimension to mass spectrometry and biomarker-oriented studies without the requirement for labels, as is the case with more traditional techniques such as fluorescense microscopy. It has particular merit for studies where no prior hypothesis of target molecules is available, as it can simultaneously track a wide range of molecules within its mass range. This makes MSI a potent exploratory tool for elucidating the spatiobiochemical topology in tissue. This paper elaborates on the principal component analysis (PCA)-based unsupervised decomposition of an MSI-measured organic tissue section into its underlying biochemical trends. We introduce a method to control the weight that particular peak intensity ranges are allowed to exert on the final decomposition model. The extension provides a way for peak intensity-based scaling to be incorporated directly into the decomposition process, for the purpose of denoising or contrast enhancement. The method makes use of peak height transformations that are conceptually equivalent to what is known in digital image processing as gray level transformations, but rather than aiming to enhance contrast for human interpretation they are used to influence the unsupervised decomposition process. As an example, we apply a combined denoising/contrast stretching measure to the MSI-measurement of a section of rat spinal cord.