Dr. rer. nat. Harald Birkholz
Member of the Research Training Group, 1.11.2008 - 31.10.2011
PhD Defence, 24.02.2012, Magna Cum Laude
Automatic quantification of cell-morphological parameters from microscopic images
The goal of this project is to identify, implement and improve suitable methods of mathematical image processing, which allow for a quantification of biological experiments by microscopic images. The automation comprises preprocessing, feature detection and feature statistics. For the particular quantification of actin filaments of fibroblasts by confocal laser scanning microscopic images we develop a software named FilaQuant for processing great numbers of images generated for the assessment of interactions between cells and biomaterials. The impact of FilaQuant is a novel approach in bright ridge detection by an oracle approach. This is combinatorially expensive but in exchange very accurate and indiferent with respect to image degradations. The accuracy should allow for high-level parameters to be derived from the data with least requirements on sample number and quality.
This project aims at support of valid research in all areas of the graduate school. Therefore data is inspected and assessed with respect to means for an automatized evaluation with great numbers of samples. The main mathematical tool is optimization with discrete models.
So far the following results have been achieved:
- A variational method for denoising image data under implicit preservation of ridges and valleys has been specified and efficiently implemented.
- A matching algorithm (PAST) for a disambiguation of superposed fibres in the plane given as a geometrically embedded graph has been formulated and efficiently implemented.
- A prototype of FilaQuant which evaluates stacks of images for length and orientation parameters of bright ridge features with optional preprocessing has been implemented using the Embarcadero RAD Studio Borland C++ Builder.
- SCHMITT, O. and BIRKHOLZ, H., 2010. A hybrid approach to quantify lamination of the cerebral cortex. Int. J. Nonlin. Sci Num, 11(5), pp. 363–370.
- SCHMITT, O. and BIRKHOLZ, H., 2011. Improvement in Cytoarchitectonic Mapping by Combining Electrodynamic Modeling With Local Orientation in High-Resolution Images of the Cerebral Cortex. Micros. Res. Tech. (74). pp. 225–243.
- BIRKHOLZ, H., 2011. A unifying approach to isotropic and anisotropic total variation denoising models. J. Comput. Appl. Math. 235 (8). pp. 2502 - 2514.
- MATSCHEGEWSKI, C., STAEHLKE, S., BIRKHOLZ, H., LANGE, R., BECK, U., ENGEL, K. and NEBE, JB.,2012. Automatic Actin Filament Quantification of Osteoblasts and their Morphometric Analysis on Microtextured Silicon-titanium Arrays. Materials, 5, pp. 1176-1195. DOI:10.3390/ma50711762012.
- BIRKHOLZ, H. and ENGEL, K.,2014. Partition into almost straight trails. Discrete Applied Mathematics (163), pp.127–135.
- BIRKHOLZ, H., MATSCHEGEWSKI, C., NEBE, J.B. and ENGEL, K., 2009. Quantification of actin filament organization by estimating graph structures in confocal microscopic images. ISBN 978-3-642-03881-5, IFMBE Proceedings, 25 (4), pp 1932-1935.
- BIRKHOLZ, H., LABAHN, R., ENGEL, K. and NEBE, J. B., 2010. Automatische Auswertung von Aktinfilament-Strukturen in konfokal-mikroskopischen Bildern. Biomedical Engineering, 55 (1), ISSN (Online) 1862-278X, ISSN (Print) 0013-5585, DOI: 10.1515/BMT.2010.553.
- VAN RIENEN, U., APPALI, R., BADER, R., BAUMANN, W., BECK, U., BEHREND, D., BENECKE, R., BIALA, K., BIRKHOLZ, H., BURKEL, E., ENGEL, G., GIMSA, J., GIMSA, U., GONGADZE, E., GRÜNBAUM, A., HABA, Y., LIESE, F., LIU, B., LÜDER, M., MATSCHEGEWSKI, C. and MITTELMEIER, W., MIX, E., NEBE, J.B., NOWAK, K.A., PAHNKE, J., PAU, H.W., PAULEWEIT, S., PETERSEN, S., POLNICK, S., REIMER, T., ROTT, G., SALOMON, R., VINTER, E., WEIHE, T., and WOLKENHAUER, O., 2010. Analysis and Simulation of Electrical Interactions of Implants with Biosystems, Biomedical Engineering. 55 (1), pp. 167–170, 2010, ISSN (Online) 1862 -278X, ISSN (Print) 0013-5585, DOI: 10.1515/BMT.2010.543.
- BIRKHOLZ, H., 2011. Extracting the Ridge Set as a Graph for Quantification of Actin Filament Images obtained by Confocal Laser Scanning Microscopy. Tenth International Conference on Quality Control by Artificial Vision, edited by Jean-Charles Pinoli, Johan Debayle, Yann Gavet, Frédéric Gruy, Claude Lambert, Proc. of SPIE, Vol. 8000, 800006. 8 pages. doi: 10.1117/12.895269.
- BIRKHOLZ, H., 2012. Extracting the ridge set as a graph for actin filament length estimation from confocal laser scanning microscopic images. Journal of Electronic Imaging, 21, (2), 021110. DOI:10.1117/1.JEI.21.2.021110.
- BIRKHOLZ, H. and ENGEL, K., 2014. Partition into almost straight trails. Discrete Applied Mathematics, 163, pp.127–135.