link to homepage

Navigation and service

Diploma dissertation: Semi-automatic segmentation of three-dimensional cortex regions of the brain

Advertising institute: ZEA-2 - Electronic Systems
Reference number: D065/2014, Engineering, computer science, physics, mathematics

The Central Institute for Electronics (ZEL) is a scientific and technical institute at Forschungszentrum Jülich GmbH. It implements research and development projects in cooperation with the institutes of Forschungszentrum Jülich and with external partners.
Work at ZEL focuses on the development of electronic and information technology system solutions in sensor and detector technology, signal and data processing, measuring technology as well as imaging and image processing.

The multimodal image processing research group at the Central Institute for Electronics focuses on investigating and developing algorithmic approaches suitable for segmenting three-dimensional structures in individual brains. This work is essential for the structural and functional analysis of these regions within the scope of neuroscientific research and clinical diagnosis and treatment. For our projects (e.g. EU, BMBF), we are therefore seeking motivated students to work on current topics in medical image processing.

Job description

A prototype for the semi-automatic segmentation of three-dimensional regions of the brain has been developed by the group on the basis of tomographic imaging data (e.g. MRI). The prototype was implemented in the MITK development environment and equipped with a user interface. It comprises a combination of 2D live-wire algorithms and the active surface model and allows the segmentation of a 3D region. The process should be flexible and compatible with 3D imaging data with different modalities.

Work on the following areas will be used to further develop and optimize this prototype:

  • Development and optimization of segmentation algorithms based on the active surface model.
  • Training to generate adaptive characteristics.
  • Expansion of the methodology for the segmentation of differently labelled brain regions.
  • Run-time optimization of the active surface model using a multi-scale approach.
  • Expansion and optimization of the graphic user interface (GUI) for use in medicine.

Phantom data will be generated for quantitative evaluation and will reflect the typical geometries of the 3D regions to be segmented. Evaluation using phantom data and real imaging data is part of quality assurance and is an important step in the process of scientific image processing.

We offer undergraduates and postgraduates an opportunity to write their diploma dissertations or master’s theses on topics in this field, which can be adapted to suit the student’s existing knowledge of the topic and the academic degree they are pursuing. Applicants should ideally have excellent programming skills (C/C++, Linux) and they should be familiar with image processing. Experience working with the software toolkits ITK,VTK, MITK and/or QT would be beneficial.

We welcome applications from students studying engineering, computer science, physics, mathematics or related area who are interested in this field, have relevant experience and work in a systematic and independent manner. Initial training will be provided.


If you would like to find out more, please contact:

Dr.-Ing. Gudrun Wagenknecht
Central Institute for Electronics
Forschungszentrum Jülich GmbH
52425 Jülich

Tel: +49 2461 61-3184
Fax: +49 2461 61-3990