Statistical methodology in cancer cell imaging
I. Ahonen, J. Nevalainen & M. Nees
Modern cancer research relies heavily on
cell-based biological models and microscopic imaging. In combination, these provide very powerful tools for studying cancer-relevant cell biology in a microenvironment that mimics the patient tumors in vitro. As a result of imaging trials, enormous amounts of data are generated that may be utilized for subsequent, detailed image analyses (morphometric or phenotypic screens). Typically, cellular differentiation or de-differentiation, or tumor-specific processes such as invasion, result in characteristic morphologies, which may also be reflected in altered gene expression patterns. Our research aims to create powerful and informative data-driven statistical modelling methods, which will allow us to analyse massive data sets that would otherwise be difficult or impossible to handle.
Our research is carried out in close collaboration with VTT Technical Research Centre of Finland, where all imaging data is generated and preprocessed.