Abstract
Deformable (2D or 3D) medical image registration is a challenging problem. Existing approaches assume that the underlying deformation is smooth. This smoothness assumption allows for solving the deformable registration at a coarse resolution and interpolate for finer resolutions. However, sliding of organs and breathing motion, exhibit discontinuities. We propose a discrete optimization approach to preserve these discontinuities. Solving continuous deformations using discrete optimization requires a fine distribution of the discrete labels. Coupled with the typical size of medical image datasets, this poses challenges to compute solutions efficiently. In this paper we present a practical, multi-scale formulation. We describe how discontinuities can be preserved, and how the optimization problem is solved. Results on synthetic 2D, and real 3D data show that we can well approximate the smoothness of continuous optimization, while accurately maintaining discontinuities.
Highlights
Deformable image registration is an important component in medical applications
We present a multi-scale Graph Cuts (GC) approach which drastically reduces the number of labels to consider at each scale
For the results in this paper we found that the Sum of Squared Differences (SSD) dissimilarity metric between moving and fixed images works well
Summary
Deformable image registration is an important component in medical applications. Examples include alignment of a patient’s MRI to a reference MRI for diagnosis, image-guided therapy and adaptive radiation therapy [15]. Individual organs typically exhibit smooth deformations, sliding motions along organ and tissue boundaries present discontinuities in the deformation fields. One approach to circumvent this discontinuity issue could be to segment each organ and perform registration on each organ separately. This requires accurate segmentation as a preprocessing step, which is not an easy task. Different organs may require different segmentation strategies This approach would not take any deformations due to organ interactions into account. We propose a deformable image registration framework based on a discrete approach with appropriate regularization which can handle local discontinuities in deformations efficiently, and accurately preserve smoothness globally. In the remainder of this paper we refer to pixels, which could mean either 2D pixels or 3D voxels
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
