Vacatures

At the moment we're looking for:

---

PhD student m/f

The position

As a PhD student in the Department of Radiology, you will work in the Neuro-Image Processing Section within the Division of Image Processing (LKEB). Your work will involve developing feature extraction approaches using novel image processing techniques with an aim to further the clinical use of MR imaging. You will work in collaboration with clinical researchers at LUMC and researchers from the National Laboratory of Pattern Recognition, CASIA, Beijing, China.

The project

MRI provides unique opportunities for the quantification of brain structures as well as the assessment of brain function. As such, MRI may be used to assess the neurological health status of a given individual based on the comparison of patient-specific MR-based features and parameters with values obtained from reference populations. Further advances in MR imaging in combination with novel automated image analysis techniques are needed to bring this technology within reach of routine clinical application.

Feature extraction approaches using novel image processing techniques will be developed along two lines: first regarding the modelling of shape and texture of multiple brain structures in a single statistical model and, second, through the comparison of functional networks. Researchers from the National Laboratory of Pattern Recognition, CASIA, Beijing, China will work on classification techniques using those features.

The department

You will work within the Division of Image Processing (LKEB) at the Department of Radiology and alongside other partners in the project. The Division investigates image analysis techniques which facilitate the interpretation of biomedical images in general. An important research field is the development of contour detection algorithms which are robust enough to perform objective quantitative measurements in medical images as they are acquired in daily clinical practice. The LKEB carries out its fundamental image processing research in a highly clinical setting as it is critically important to ascertain clinical input from different disciplines, such as cardiology, radiology, pulmonology, surgery and orthopaedics.

Your profile

You hold an MSc degree in a relevant discipline such as Computer Science or Electrical and Computer Engineering, and have formal training in signal and image processing. You should have an affinity for biomedical research with a clear interest in the application and utilisation of biomedical and medical technological research. An extensive background in programming and experience in medical image processing is essential, as is knowledge of Linux, Windows and Matlab. You should have had at least one scientific paper published in an international forum. In addition, you are a committed and enthusiastic team player with excellent communication skills. You are attentive to detail and possess an ability to solve problems at an abstract level yet able to pragmatically apply these solutions in practice.

We offer

You will be employed on the basis of a 36-hour week. Appointment is for an initial period of 1 year, possibly to be extended by a further 3 years conditional on a positive first-year appraisal. Your salary will depend on your qualifications and experience, with a maximum of € 3,899 gross per month based on a full time position (scale 10 of the Collective Labour Agreement for University Hospitals).

More information

If you have any questions, or if you want more information about this position, please contact Professor B.P.F. Lelieveldt, Department of Radiology, e-mail B.P.F.Lelieveldt@lumc.nl, telephone +31 71 526 3935 or 526 1882 or Dr Julien Milles, assistant professor, Department of Radiology, e-mail J.R.Milles@lumc.nl, telephone +31 71 526 5342.

e-mail: DIV2PenO@lumc.nl

Applying by letter:

LUMC
Attn. mrs drs. P. Vis, P&O advisor
Bureau Bedrijfsvoering Division 2 (K5-S)
PO Box 9600
2300 RC Leiden
The Netherlands

---

For Chinese Graduate Students only:

We are looking for excellent Chinese graduate students, for PhD projects to be funded by the China Scholarship Council (CSC).
Interested, motivated and top-level graduate students are encouraged to contact us. Upon agreement we can create a research proposal for the 2012 CSC funding round.

The LKEB is part of the Leiden University Medical Center (LUMC), The Netherlands, and is thus part of Leiden University. Leiden University was the first university of The Netherlands (founded 1575) and has a renowned history with 15 Nobel prizes among our alumni and various famous researchers like Huygens, Descartes, Lorentz and Einstein. One of the pillars of the university has always been Medicine, which is nowadays focussed in the LUMC. A trend in medicine is the expanding role of imaging, requiring the interdisciplinary cooperation between medical doctors and technical scientists. Our department takes a position right there. The group’s main goal is the development and validation of innovative automated techniques (segmentation, registration, classification, visualisation, quantification) for the analysis and processing of medical and biological images. In addition, we aim to make high quality software, for daily use by clinicians, which is accommodated by the spin-off company Medis medical imaging systems. Being embedded in the Department of Radiology, LKEB benefits from a unique collaboration with clinicians and access to clinical data, which is critical for the success of medical image processing research. We offer an excellent multi-disciplinary research environment in an internationally recognized group.

We are looking for excellent CSC fundable students with a master in a technical science, such as physics, electrical engineering, mathematics, technical medicine, or similar. Successful candidates possess very good English communication skills (reading, writing and speaking), and have passed their Master exams with high marks. Knowledge of image processing, and/or programming is considered an advantage. (see also pdf ducument at the end of this page)

Possible PhD projects

Together with the candidate we can define a very interesting and cutting edge research proposal, to be submitted for the CSC 2012 grant round. The exact topic and content of the proposal will depend on the student’s background and interest, and on the research directions currently pursued within the LKEB. Below we give five example project descriptions, that can be worked out to a full research proposal.

• Project 1: Fast Image Registration for Time-critical Medical Applications.
  In the modern operating room intra-operatively acquired imaging data is used amongst others for surgery guidance and radiotherapy. The processing of these data usually requires a matching or image registration step, and needs to be near real-time, since doctors have to act quickly. Current registration algorithms are too slow for these procedures. Therefore in this project we aim for increasing their performance by means of parallel processing techniques. New (mathematical) tools for the parallel optimization of image registration need to be developed, substantially advancing the aforementioned clinical procedures.
  For more information contact: Marius Staring PhD, m.staring@lumc.nl
• Project 2: Localized Multi-Atlas-Based Segmentation.
  The segmentation of structures like organs or bones from 3D images of the human body is one of the biggest problems in medical image processing. A widely-used and successful approach is that of using multiple atlases, which after image registration are fused to a single segmentation. In this research we will study and further develop the fusion rule as a classifier fusion problem. Incorporation of local fusion will be an important aspect of the research, since the image registration quality (base classifier quality) varies locally. The leading application will the segmentation of lung lobes from chest CT data, but extension to other data is very interesting.
  For more information contact: Marius Staring PhD, m.staring@lumc.nl
• Project 3: Tissue Characterization and Multi-Modality Fusion in Optical Coherence Tomography Images.
  Optical Coherence Tomography is a new imaging modality which provides real-time high resolution images of the coronary vessels. Besides providing detailed information about stent appositions and lumen dimensions, it also provides information about the vessel wall composition, like fibrotic caps and necrotic cores. The interpretation is still quite challenging and therefore new tissue characterization techniques should be developed and tested to come to a better understanding of the OCT images. The information should be fused with other available images from the same vessel, e.g. CT or IVUS. Besides coronary arteries, also the feasibility of using OCT in other arteries will be studied.
  For more information contact: Jouke Dijkstra PhD, j.dijkstra@lumc.nl
• Project 4: Classification-Based Screening of Very Large Data Cohorts.
  The brain is a deceiving organ: despite its debonair appearance, its functioning is highly complex and not fully understood yet. For example, neurological ailments are very difficult to be related to a single cause. As a result, image processing applied to brain images is gearing towards the use of image-based features in classification frameworks to allow the identification of a particular disease. When considering very large cohorts, group identification becomes difficult if not impossible. This research project will tackle that issue by making use of one-class classifiers, where deviation from normal do not need to be specified on beforehand. The main focus of this research will be neuro-radiological MR data, but the framework is expected to be applicable to other modalities and organs.
  For more information contact: Julien Milles PhD, j.r.milles@lumc.nl
• Project 5: Multi-modality imaging and quantitative assessment of coronary bifurcations
  Coronary artery disease is one of the leading causes of death worldwide. The continuous drive for optimal patient care demands intuitive visualization of coronary vascular structures as well as accurate and reproducible quantifications. Bifurcation lesions are frequent and its treatment remains challenging in current routine clinical practice. High-quality imaging and quantification of coronary bifurcations is crucial for the understanding of coronary anatomy and the response to its therapeutics. While X-ray coronary angiography provides an excellent global overview of coronary vascular structures, novel intracoronary imaging techniques such as intravascular ultrasound (IVUS) and optical coherence tomography (OCT) documents detailed information in the vessel wall including plaque composition and lesion extent. The fusion of the two imaging modalities provides complementary information for the assessment of bifurcation anatomy and for support of bifurcation interventions. In this project, a real-time multi-modality image fusion approach has to be developed and validated extensively in typical clinical population. New approaches for the quantification of bifurcation lesions will also be created. This project is in close collaboration with Medis medical imaging systems bv (www.medis.nl) and several key opinion leaders in interventional cardiology. For more information contact: Shengxian Tu PhD, S.T.Tu@lumc.nl, or Prof. Johan H. C. Reiber PhD, J.H.C.Reiber@lumc.nl.

Note that available research topics are not limited to the examples above. It is expected that the PhD research will result in four publications in the top 20% journals of our field. In addition, several international conferences will be visited.

Interested student are encouraged to contact the director of the LKEB, prof. Lelieveldt, b.p.f.lelieveldt@lumc.nl, or the contact person of one of the projects directly.

Downloads

• CSC Students Information (PDF, 605.7 KB)