Related Experiment Video
Updated: Jul 10, 2026
09:11
Use of MRI-ultrasound Fusion to Achieve Targeted Prostate Biopsy
Published on: April 9, 2019
3D prostate boundary segmentation from ultrasound images using 2D active shape models
1Department of Medical Biophysics, The University of Western Ontario, London, Canada. ahodge3@uwo.ca
Summary
This study presents a new semi-automatic algorithm for prostate segmentation in 3D ultrasound images. The method achieves accurate boundary outlining with improved speed, aiding cancer diagnosis and treatment planning.
Area of Science:
- Medical imaging
- Biomedical engineering
- Computational anatomy
Background:
- Accurate prostate segmentation is crucial for effective prostate cancer diagnosis and treatment planning.
- Current manual methods can be time-consuming and prone to inter-observer variability.
Purpose of the Study:
- To develop and evaluate a semi-automatic algorithm for three-dimensional (3D) prostate boundary segmentation using ultrasound images.
- To assess the accuracy and efficiency of the proposed segmentation method compared to manual outlining.
Main Methods:
- The algorithm utilizes two-dimensional (2D) active shape models (ASM) combined with rotation-based slicing for semi-automatic 3D segmentation.
- Evaluation involved comparing algorithm-generated boundaries against manually delineated gold standards using distance and volume metrics.
Main Results:
- The algorithm achieved a mean absolute distance of 1.09+/-0.49 mm between segmented and gold standard boundaries.
- An average absolute volume difference of 3.28+/-3.16% was observed.
- A 5x speed increase was demonstrated compared to manual planimetry.
Conclusions:
- The developed semi-automatic segmentation algorithm provides accurate and efficient prostate boundary outlining from ultrasound data.
- This method holds potential for improving the workflow in prostate cancer diagnosis and treatment planning.
