Research Group 2
Stochastic Modelling and Statistics
This research unit in the Institute of Biomathematics and Biometry at the HelmholtzCenter Munich is concerned with probabilistic and statistical modelling, classical statistics, as well as statistical analysis and exploration of data. Emphasis is on biological and medical applications.
The group has expertise in mathematics, statistics, as well as time series and image analysis. It consists of scientists with mathematical, statistical and engineering background. There are intimate connections to the analysis, numerical, and computational groups in the IBB, and also to the experimental institutes at the HelmholtzCenter Munich, in particular those concerned with genetics, molecular- and microbiology, pathology and imaging.
The group covers a fairly broad scope of scientific topics and interests. Main research areas can be specified as follows:
Analysis of One- and Multi-Dimensional Signals
for example time series, multi-dimensional data sets like images or series or stacks of images. In this part of our research, we focus on the extraction of morphological and spatial rather than quantitative features. Emphasis is on parsimonious and sparse data representations, based on variational approaches with complexity penalized likelihood functions. Real time algorithms for the solution of the respective optimization problems are a main part of this kind of research.
Another topic are discontinuities in digital images. Focus is on the design and on the theoretical study of geometry-based non-linear approximation methods for the representation of edges and lines in images. The main approaches are partition-regression (quadtree structures, wedgelets) and Delaunay triangulations.
There is also a body of work on probability theory and statistics in their own right.
some signals analysis publications
Mathematical Systems Biology
in particular probabilistic modelling of autoregulative intra- and intercellular mechanisms, as well as consistency in the limit of large numbers of molecules in certain compartments, while other types of molecules are rare. Typical examples are cellular clocks, appearing for instance in somitogenesis in the context of members of the Hes-family.
some systemsbiology publications
We are concerned with the investigation and exploration of data, as well as descriptive and inferential data analysis.
some biostatistical publications
Analysis of (Bio-) Medical Image Data
in the context of magnetic resonance imaging (MRI) and microscopy. Special topics are brain mapping, diffusion tensor imaging, functional MRI, as well as the morphological characterization of brain structures in order to discriminate healthy and morbid tissue. Main goal is to support diagnosis and therapy of diseases in the context of aging, like the ubiquitous Alzheimer disease.
Additionally the quantitative representation and analysis of entities in data from fluorescence microscopy and confocal laser scanning microscopy (CLSM) mainly in histopathology and microbiology (oncology, microbial growth, biofilm, rhizosphere) are dealt with.
some methodical publications
is a consulting and service unit intended to serve as an interface between external scientists and the diverse IBB groups. It provides assistance for the solution of practical problems, in particular in statistics or image analysis, but also in other fields of mathematics.
ImStatLab Hotline: e-mail