Mass spectrometry imaging (MSI), in particular matrix-assisted laser desorption/ ionization time-of-flight (MALDI-TOF), is a label-free technique for spatially resolved molecular analysis of tissue samples with a broad range of applications in life sciences and biomedical research. It has become a routine technique for applications in drug discovery, biomarker detection or several Omics-methods. MSI, however produces extremely completely data sets, which cannot be evaluated without bioinformatics tools.
The success of MSI data analyses is based on a combination of technological hardware developments and the development of novel mathematical theories for extracting meaningful information. Data sets contain mass spectra, which are measured at different locations (pixels) of a tissue section. This yields a hyperspectral data set with thousands of channels (mass-to-charge values) at every pixel.
The mathematical algorithms used for MSI data analyses are based on theoretical research at the interface of functional analysis (inverse problems) and numerical linear algebra (matrix factorization). One important task is to determine characteristic spectral patterns in the data, which are the building blocks for classification schemes (e.g. in tumor typing) or identification (e.g. for biomarker detection).
The challenges and methods described are the core of the research activities of the Center for Industrial Mathematics (ZeTeM), University of Bremen, Germany. ZeTeM provided the mathematical foundations as well as – in collaboration with the national leading pathological service institute Proteopath GmbH in Trier, Germany – prototypical implementations for supporting pathological diagnosis. The resulting publications and patents were the basis for founding the spin-off company SCiLS GmbH, which developed a commercial software for analysing MSI data. This software has a market share of 80% worldwide and was acquired by Bruker Daltonics, the worldwide market leader in mass spectrometry imaging.