Related Experiment Videos
Optimal femoral head contour segmentation in CT images using dynamic programming.
Gaëtan Marti1, Charles Baur, Pierre-Yves Zambelli
1IPR-LSRO-VRAI Group, Ecole Polytechnique Fédérale de Lausanne, CH-1015 Lausanne, Switzerland. gaetan.marti@epfl.ch
Summary
This study introduces a novel active ray contour segmentation method for detecting the femoral head in CT scans, offering efficient and automatic segmentation even for pathological bone structures.
Area of Science:
- Medical Imaging
- Computer-Aided Surgery
- Orthopedics
Background:
- Accurate segmentation of the femoral head in CT images is crucial for diagnosing hip conditions and surgical planning.
- Conventional methods often struggle with pathological bone structures and require iterative processes.
Purpose of the Study:
- To present a new, efficient, and automatic contour segmentation method for femoral head detection in CT images.
- To enable the creation of accurate 3D hip models for preoperative planning.
Main Methods:
- Utilizing active rays to represent the femoral head contour.
- Incorporating internal energy (continuity, smoothness) and external energy (image edges, prior knowledge) for contour definition.
- Employing dynamic programming for a globally optimal, non-iterative solution.
Main Results:
- The proposed method achieves automatic segmentation of the femoral head, including pathological bone structures.
- Demonstrated higher efficiency compared to conventional methods in clinical pelvis CT scans with degenerative hip changes.
- The approach facilitates the generation of graphical 3D hip models.
Conclusions:
- The active ray contour segmentation method provides an efficient and accurate solution for femoral head detection in CT imaging.
- This technique holds significant potential for improving preoperative planning in orthopedic surgery.
- The method's ability to handle complex bone structures enhances its clinical applicability.