Related Experiment Video
Updated: May 4, 2026
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
3.6K
Prostate Segmentation in CT Images via Spatial-Constrained Transductive Lasso.
Yinghuan Shi1, Shu Liao, Yaozong Gao
1State Key Laboratory for Novel Software Technology, Nanjing University, China ; Department of Radiology and BRIC, UNC Chapel Hill, U.S.
Proceedings. IEEE Computer Society Conference on Computer Vision and Pattern Recognition
|December 17, 2013
Summary
This study introduces a semi-automated method for prostate segmentation in CT images, improving accuracy for image-guided radiotherapy. The novel approach requires minimal physician input, enhancing clinical feasibility.
Area of Science:
- Medical Imaging
- Radiotherapy
- Computational Anatomy
Background:
- Accurate prostate segmentation is crucial for effective image-guided radiotherapy.
- Current segmentation methods face challenges in precision and efficiency.
Purpose of the Study:
- To develop a novel semi-automated method for accurate prostate segmentation in CT images.
- To improve the efficiency and clinical feasibility of prostate segmentation for radiotherapy.
Main Methods:
- A two-step semi-automated approach involving prostate-likelihood estimation using Spatial-COnstrained Transductive LassO (SCOTO).
- Multi-atlases based label fusion utilizing shape information from planning and previous treatment images.
- Minimal manual input required from physicians, specifying only the first and last prostate slices.
Main Results:
- The proposed method demonstrated superior performance compared to state-of-the-art techniques on a real prostate CT dataset (24 patients, 330 images).
- Achieved accurate prostate segmentation with minimal physician interaction time.
Conclusions:
- The novel semi-automated method offers a clinically feasible and accurate solution for prostate segmentation in CT images.
- This technique enhances image-guided radiotherapy by improving segmentation precision and efficiency.
