Related Experiment Video
Updated: Jun 21, 2026
05:05
Four-Dimensional CT Analysis Using Sequential 3D-3D Registration
Published on: November 23, 2019
Integrated segmentation and non-linear registration for organ segmentation and motion field estimation in 4D CT data
A Schmidt-Richberg1, H Handels, J Ehrhardt
1Department of Medical Informatics, University Medical Center Hamburg-Eppendorf, Hamburg, Germany. a.schmidt-richberg@uke.uni-hamburg.de
Methods of Information in Medicine
|July 8, 2009
Summary
This study introduces an integrated variational method for simultaneous segmentation and motion estimation in medical imaging. The approach improves accuracy for applications like adaptive radiation therapy compared to independent methods.
Area of Science:
- Medical Imaging
- Computational Anatomy
- Radiotherapy Physics
Background:
- Spatiotemporal tomographic imaging generates large datasets, necessitating automated segmentation and motion estimation.
- Adaptive radiation therapy requires accurate estimation of organ motion, particularly respiration-induced motion, for effective treatment.
Purpose of the Study:
- To present a variational approach for simultaneous segmentation and dense non-linear registration of 3D image sequences.
- To improve segmentation and registration quality by integrating mutual prior information.
Main Methods:
- Combines variational region-based level set segmentation with diffusive registration of spatial images.
- Introduces a novel energy term to integrate segmentation and registration processes.
Main Results:
- Successfully applied to segment the liver and estimate respiration-induced motion from 4D thoracic CT images.
- Demonstrated improved segmentation and motion estimation results compared to conventional uncoupled methods.
Conclusions:
- The integrated approach is beneficial for simultaneous segmentation and registration in thoracic tumor radiation therapy.
- This method enhances results compared to applying segmentation and registration techniques independently.
