Deformable Registration

Research interests in single-modality nonlinear registration include four different kinds of subproblems. Deformable registration of two or more CT lung data sets at different states in the breathing cycle going from Functional Residual Capacity (FRC, expiration) to Total Lung Capacity (TLC, inspiration) for modelling breathing motion and deriving lung ventilation. Deformable registration of a contrast-enhanced and a native CT lung data set for deriving lung perfusion. Deformable registration of contrast-enhanced and native CT liver data sets at one or several phases in the contrast-uptake cycle for liver perfusion. And finally, highly accurate partially rigid bone registration for head and neck CT-Angiography applications to extract bone structures from CTA images.


Martin Urschler, see also personal ICG website.
This topic was mainly dealt with by Martin Urschler during his PhD thesis, with additional help from Thomas Pock and later Manuel Werlberger who applied TV based regularization priors to optimize energy minimization models for deformable registration. Related master student works:
  • Sasa Grbic did his master's thesis on nonlinear registration using a second order TV based regularization prior.
  • Stefan Kluckner did his master's thesis developing a framework for evaluation of nonlinear registration methods.

Funding Source

Record not found: ResearchProject

Project Related Publications

Robust Optical Flow Based Deformable Registration of Thoracic CT Images We present an optical flow deformable registration method which is based on robust measures for data and regularization terms. We
show two specific implementations of the method, where one penalizes gradients in the displacement field in an isotropic fashion and the other one regularizes by weighting the penalization according to the image gradients anisotropically.
Urschler M, Werlberger M, Scheurer E, Bischof H: Presented at MICCAI Workshop Medical Image Analysis in the Clinic: A Grand Challenge 2010, Beijing, China. PDF
Optical flow based deformable volume registration using a novel second-order regularization prior
Nonlinear image registration is an initial step for a large number of medical image analysis applications. Optical flow based intensity registration is often used for dealing with intra-modality applications involving motion differences. In this work we present an energy functional which uses a novel, second-order regularization prior of the displacement field.
Grbic S, Urschler M, Pock T, Bischof H: Presented at SPIE Symposium on Medical Imaging: Image Processing (SPIE:MI) 2010, San Diego, USA. PDF
A Framework for Comparison and Evaluation of Nonlinear Intra-Subject Image Registration Algorithms A framework for evaluation of nonlinear registration algorithms is presented in this work. We give an open-source implementation based on ITK algorithms and provide means for extending the framework in the future. Urschler M, Kluckner S, Bischof H: Presented at MICCAI Open Science Workshop 2007, Brisbane, Australia. PDF
A Duality Based Algorithm for TV-L1-Optical-Flow Image Registration Nonlinear image registration is a challenging task in the field of medical image analysis. In many applications discontinuities may be present in the displacement field, and intensity variations may occur. In this work we therefore utilize an energy functional which is based on Total Variation regularization and a robust data term. We show experimental results on synthetic and clinical CT lung data sets at different breathing states as well as registration results on inter-subject brain MRIs. Pock T, Urschler M, Zach C, Beichel R, Bischof H: Presented at International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) 2007, Brisbane, Australia. PDF
Automatic Point Landmark Matching for Regularizing Nonlinear Intensity Registration: Application to Thoracic CT Images A deformable registration algorithm that combines feature-based (SIFT and shape context matching) and intensity-based optical flow methods into a common framework is shown here. Application to thoracic CT data shows its practical relevance. Urschler M, Zach C, Ditt H, Bischof H: Presented at International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) 2006, Copenhagen, Denmark. PDF
A new registration/visualization paradigm for CT-Fluoroscopy guided RF liver ablation Joint work with Ruxandra Micu from Siemens and the lab of Prof. Navab in Munich. The topic of this work is 2D-3D registration in ablation therapy applications. Micu R, Jakobs T, Urschler M, Navab N: Presented at International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) 2006, Copenhagen, Denmark. PDF
SIFT and Shape Context for Feature-Based Nonlinear Registration of Thoracic CT Images In this work a feature based registration pipeline for computing nonlinear deformations in intra-modality CT applications is presented. It is based on SIFT and shape context and applied to thoracic data. Urschler M, Bauer J, Ditt H, Bischof H: Presented at European Conference on Computer Vision Workshop: Computer Vision Applications in Medical Image Analysis (CVAMIA) 2006, Graz, Austria. PDF
Partially rigid bone registration in CT Angiography Here we present a partially rigid bone registration algorithm to improve visualization of MIP CT Angiography studies where several independent rigid motions occur during acquisition of the two images (with and without contrast agent) which are used for subtraction. The algorithm improves the Matched Mask Bone Elimination technique. Urschler M, Ditt H, Bischof H: Presented at Computer Vision Winter Workshop 2006, Telc, CZ. PDF
Assessing breathing motion by shape matching of lung and diaphragm surfaces In this work we show a surface based nonlinear registration algorithm using the 3D shape context approach applied to lung, airway tree and diaphragm surfaces to assess breathing motion. Correspondences are found from 3D shape context matching and the registration makes use of a thin-plate-spline displacement field interpolation. Here we focus on the evaluation of the registration algorithm./p> Urschler M, Bischof H: Presented at SPIE Symposium on Medical Imaging 2005: Physiology, Function, and Structure from Medical Images (SPIE) 2005, San Diego, US. PDF
Registering 3D lung surfaces using the shape context approach Here we investigate the 3D shape context approach for registration of medical images. Surfaces, in our case from a lung segmentation, are extracted from the data and the shape context matching is combined with a thin-plate-spline registration approach to get a dense displacement field. Urschler M, Bischof H: Presented at Eighth Annual Conference Medical Image Understanding and Analysis (MIUA) 2004, London, UK. PDF
Matching 3D lung surfaces with the shape context approach In this work we investigate the extension of the 2D shape context to 3D surfaces for matching of segmented medical structures. As an example we show segmented lung surfaces, derive surface points, calculate the descriptors and match the surfaces. Urschler M, Bischof H: Presented at 28th Workshop of the Austrian Association for Pattern Recognition (OAGM) 2004. PDF
Nonlinear intra-modality registration of medical volume data Urschler M: PhD Thesis. PDF