Division of Medical Image Computing

PD Dr. Klaus Maier-Hein

Magnetic resonance imaging produces a wealth of information which we combine in personally adapted computational models of living organs. This image shows a graph theory-based model of the human brain that helps us learn, detect, and predict disease patterns.
© dkfz.de

Medical images uniquely represent the anatomical and functional progress of diseases in 3D space and time. The Division of Medical Image Computing strives to utilize the vast and unexploited potential in these images through computational image understanding and information processing.

“Radiomics” denotes the emerging endeavor of systematic extraction, mining and leveraging of this rich information towards personalized medicine. We aim to comprehensively summarize imaging information from multiple time-points and modalities in condensed, quantitative signatures and link them with clinical and biological parameters (e.g. genomics or proteomics). We develop our methods for various clinical applications, with a particular emphasis on prostate cancer, breast cancer and brain tumors.

Another focus of our research lies in processing, analysis and visualization of neurological datasets, especially from diffusion-weighted MRI. We develop techniques for white matter fiber tractography and segmentation, as well as for brain connectivity analysis (connectomics). Main fields of application comprise Alzheimer’s disease, autism spectrum disorder and borderline personality disorder, as well as surgery planning and navigation in the context of tumor treatment.


The applications mentioned above require cutting edge developments at the core of computer science. We have a profound track record in the methodology of machine learning, especially in the context of big data applications where large-scale heterogeneous data sources are analyzed. Transparent deep learning techniques that are interpretable and that explicitly deal with uncertainty in the data are of particular interest to us. We further pursue novel image computing concepts that combine mathematical modelling approaches with current machine learning technology. They can enable the simultaneous optimization of all required components in end-to-end training scenarios.

We put a strong focus on the successful validation and translation of the developed techniques into clinical practice. To this end, a dedicated group for scientific software engineering was established, which coordinates and implements IT strategies of the research program “Imaging and Radiooncology” at DKFZ and contributes to several consortia such as the German Cancer Consortium (DKTK) or the National Center for Tumor Diseases (NCT). Our versatile and open technological portfolio builds a foundation for national and international projects and is constantly advanced to fulfill the needs of current medical imaging research.


PD Dr. Klaus Maier-Hein
Medical Image Computing (E230)
Deutsches Krebsforschungszentrum
Im Neuenheimer Feld 581
69120 Heidelberg
Tel: +49 6221 42 3545

Selected Publications

  • Maier-Hein KH, Neher P, Houde JC, Cote MA, Garyfallidis E, Zhong J, Chamberland M, et al. “The Challenge of Mapping the Human Connectome Based on Diffusion Tractography.” Nature Communications, accepted 2017.
  • Norajitra T, Maier-Hein KH. “3D Statistical Shape Models Incorporating Landmark-Wise Random Regression Forests for Omni-Directional Landmark Detection”. IEEE Transactions on Medical Imaging, 36(1):155-168, 2017.
  • Neher, PF, Côté MA, Houde JC, Descoteaux M, Maier-Hein KH. “Fiber Tractography Using Machine Learning”. NeuroImage, 158: 417–29, 2017.
  • Nolden M, Zelzer S, Seitel A, Wald D, Müller M, Franz AM, Maleike D, Fangerau M, Baumhauer M, Maier-Hein L, Maier-Hein KH, Meinzer HP, Wolf I: The Medical Imaging Interaction Toolkit: challenges and advances. International Journal of Computer Assisted Radiology and Surgery, 1-14, 2013.
to top