Automated segmentation and contour propagation

Host: Faculty of Mathematics and Physics of the University of Ljubljana (UL), Slovenia.

Planned secondments: University of Aarhus (Denmark), PSI (Switzerland) and University of Wisconsin (USA).

Project description

One of the key aspects of the daily on-line adaptive workflow is to guarantee fast and accurate contours on the daily CT. These could be obtained in three ways: i) by propagating the planning contours to the daily CT pre-registered to the planning CT; ii) by performing a daily (automatic) segmentation (i.e. considering the daily CT as an independent 3D image), or iii) a combination of these approaches (e.g., by first propagating approximate contours and then improving them based on daily image). Auto contouring applies both to tumours as well as critical tissues/normal organs. Both approaches are in clinical use, in most of the cases, however, only as helping tool to speed up the contouring process. Indeed, typically, the clinician needs to refine (some) of the contours before these can be used clinically. The step of the manual fine-tuning has however to be skipped (or reduced) in an on-line adaptive workflow, to guarantee a fast and efficient workflow.

It is still unclear what are errors associated with these two methods. Therefore the studies will be divided among development of suitable automatic segmentation tools and methodology for objective assessment of residual errors, particular in terms of associated impact on treatment both with respect to tumour control as well as associated adverse effects. In scope of the RAPTOR as a whole, the focus will be on daily adaptation of treatment, and hence propagation of tumour and organ contours based on daily-recorded anatomical images; however, strategies for more rapid (near real time) will also be considered. The methodology will be tested with different clinical indications and the quality of the achieved daily contour will be systematically compared to reference contours.

The clinical impact of the residual errors will also be assessed for the different indications. The goal of the project is to indicate which modality has to be preferred for the on-line adaptive approach and possibly to suggest the smaller error in the daily contours that can be clinically accepted.

The PhD student will be will be enrolled in the PhD program at the Faculty of Mathematics and Physics at the University of Ljubljana (UL), will be a member of the Medical Physics Research Program ( and will be supervised by Dr. Andrej Studen. UL is the largest university of Slovenia and hosts multiple faculties dedicated to research and study areas. The Medical Physics Research Program provides a comprehensive environment for studying medical physics-related fields, communication between clinical and research personnel, and translation of physics tools to clinical practice.

For more information concerning the research project please contact: Andrej Stunden

Name Surname

Job title

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

University of Ljubljana -The Faculty of Mathematics and Physics


The Faculty of Mathematics and Physics at the University of Ljubljana is home to the Department of Physics and Department of Mathematics. Researchers of the faculty play a leading role in major international scientific collaborations, e.g. the Belle collaboration for research in the physics of elementary particles in Tsukubi, Japan, the Large Hadron Collider in Cern, Switzerland and the Thomas Jefferson National Accelerator Facility in Newport News, Virginia, USA.