Computational Patient Models

Image processing and patient modelling group in radiation oncology

headed by Dr. Kristina Giske

 

We are physicists, medical informaticians, computer scientists, mathematicians, and radiooncologists working together for the benefit of the patient.

You also want to join our team within your thesis project? Get in touch

Our research is dedicated to computer-aided simulation of the cancer patient receiving a curative radiotherapy treatment. The challenge for technically precise treatment techniques is the adaptation of dose delivery to motion- and physiologically induced deformations in the patient’s anatomy. Our vision is to design a virtual patient model, which can – once aligned with each individual patient – predict the therapeutic effect of the planned treatment. Thus, allowing the treating physicians to tailor the treatment to the specific needs of the patient. Algorithm development, emerging computer technology, and our passion for unusual solutions are our tools in the fight against cancer.

Static in-silico patient models become alive through bio-mechanical motion laws. Motion vector fields help to monitor and optimize the accumulating dose distribution within each individual patient.

Research focus

The strength of radiation therapy - as curative cancer treatment - is to precisely shape the dose deposition during delivery to a locally restricted target volume sparing healthy organs. As consequence, radiotherapy is capable to reduce toxicity and adverse effects while simultaneously maintaining the efficiency of tumor control. Based on tomographic imaging of the patient’s anatomy a virtual patient model needs to be constructed inside a computer program to guide the targeted planning process aligning the coordinate systems of the radiation beam and the target volume inside the patient. Without adapting the beam position and shape inevitable patient deformation due to respiratory motion, pose change, and weight loss would result in deviations of the optimized treatment. Here is where our expertise comes into play!

Our projects aim at teaching the computer model

  • …to breath like the patient by detecting the respiratory motion patterns applying efficient image processing methods
  • …to change pose like the patient by bio-mechanical motion modelling using the degrees of freedom of body structures
  • …to imitate the variability of patient’s organs caused by e.g. the bladder filling applying bio-mechanically controlled constrains to the organ’s neighborhood
  • …to lose weight or swallow like the patient by controlling the heterogeneity parameters of the involved soft tissue types

Controlling the virtual model aligned to the individual patient using our methods will allow us

  • …to identify the possible risks to fall behind the therapeutic goals  
  • …to develop strategies to compensate for the identified risks
  • …to guide the offline and online adaptation process of the dose delivery

Current projects

  • Biomechanical patient modelling
  • GPGPU parallelisation for image processing algorithms
  • OpenGL/CUDA visualization techniques
  • Virtual Reality head-mounted device applications
  • Deep learning for tissue segmentation using neuronal networks
  • Voxelized, tessellated, and analytical patient representation
  • Handling different MRI scans for MR-guided treatment

People

  • Dr. Kristina Giske (Group leader)
  • Markus Stoll (PhD student)
  • Hendrik Teske (PhD student)
  • Kathrin Bartelheimer (PhD student)
  • Angelika Czekalla (master's student)
  • Henry Müssemann (bachelor's student DHBW)
  • Dr. Eva Stoiber (radiation oncologist)

Alumni:

  • Paul Mercea (PhD student)
  • Julian Suleder (master's student)
  • Simon Kirchhof (master's student)
  • Luis Fernando Paredes Ocampo (master's student)
  • Johannes Merz (master's student)
  • Daniel Schaubach (master's student)
  • Thomas Wollmann (scholarship student)
  • Angelika Laha (bachelor's student)
  • Jan Meis (student assistant)
  • Anna Storz (student assistant)
  • Nico Schweiger (student assistant)
  • Sarah Grimm (student assistant)

Funded by

we gratefully acknowledge our supporting organisations

Selected publications

Stoiber EM, Bougatf N, Teske H, Bierstedt C, Oetzel D, Debus J, Bendl R, Giske K. 2017 Analyzing human decisions in IGRT of head-and-neck cancer pateints to teach image registration algorithms what experts know. Radiat Oncol. 12:104

Teske H, Bartelheimer K, Meis J, Bendl R, Stoiber EM, Giske K. 2017 Construction of a biomechanical head and neck motion model as a guide to evaluation of deformable image registration. Phys Med Biol. 62(12):N271-N284

Teske H, Bartelheimer K, Bendl R, Stoiber EM, Giske K. 2017 Handling images of patien postures in arms up and arms down position using a biomechanical skeleton model. Current Directions in Biomedical Engineering 3(2):469-472

Bartelheimer K, Teske H, Bendl R, Giske K. 2017 Tissue-specific transformation model for CT-images. Current Directions in Biomedical Engineering 3(2):525-528

Stoll M, Stoiber EM, Grimm S, Debus J, Bendl R, Giske K. 2016 Comparison of Safety Margin Generation Concepts in Image Guided Radiotherapy to Account for Daily Head and Neck Pose Variations. PLoS One 11(12):e0168916

Teske H, Mercea P, Schwarz M, Nicolay NH, Sterzing F, Bendl R. 2015 Real-time markerless lung tumor tracking in fluoroscopic video: Handling overlapping projected structures. Med Phys. 42(5):2540-9

Stoll M, Giske K, Debus J, Bendl R, Stoiber EM. 2014 The frequency of re-planning and its variability dependent on the modification of the re-planning criteria and IGRT correction strategy in head and neck IMRT. Radiat Oncol. 9:175

 

 

to top