We focus on automated image and data analysis with respect to biological, biochemical, and clinical application fields under the “Biointerfaces in Technology and Medicine (BIFTM)” program of the Helmholtz Association. Our work mainly deals with high-throughput methods and the analysis of image data from high-resolution microscopy.
In addition, we also analyze data from energy systems, some of which are part of the Energy Lab 2.0 and Energy Systems 2050 projects.
Most problems come about from deriving measures and making decisions based on large, partially redundant and strongly damaged datasets. Here, a purely manual data analysis is often overwhelming to users due to the many different attributes and complex correlations.
Our projects aim to develop intelligent methods, such as
- fuzzy systems,
- artificial neural networks,
- support vector machines,
- statistical methods,
- and feature extraction methods (e.g., spectral analyses, wavelet analyses, correlation analyses),
and use them for automatic data evaluation. We also enhance and refine these techniques and adapt them to specific requirements if necessary. The focus in the data evaluation is not only on automated classification but also on a comprehensible and interpretable solution approach that led to the specific decisions. This ensures rationality of the decisions and raises the acceptance level among users.
In many of our previous projects, we have managed to develop a standard basis that can be adapted to different biological, chemical, medical, and technical problems (see image). Most elements of this strategy were implemented in the MATLAB toolbox, SciXMiner, which is freely available under the GNU-GPL. In addition, we also develop the XPIWIT image analysis software for data volumes in the terabyte range.
apl. Prof. Dr. Ralf Mikut
Jorge Angel Gonzalez Ordiano
Nicole Ludwig (IPD)
PD Dr. Markus Reischl
apl. Prof. Dr. Christian Pylatiuk
Dr. Stefan Schulz
Dr. Johannes Stegmaier
Marian Turowski (TM)
- Automated Image and Data Analysis
- 3D+t Terabyte Image Analysis
- Imaging of the corneal sub-basal nerve plexus
- Energie System 2050
- EnergyLab 2.0