Related Experiment Video
Updated: Jul 13, 2026
08:12
A Volumetric Method for Quantification of Cerebral Vasospasm in a Murine Model of Subarachnoid Hemorrhage
Published on: July 28, 2018
Segmentation and quantification of human vessels using a 3-D cylindrical intensity model.
1Department of Bioinformatics and Functional Genomics, Biomedical Computer Vision Group, BIOQUANT, and IPMB, University of Heidelberg, D-69120 Heidelberg, Germany. s.woerz@dkfz.de
Summary
This study presents a novel 3-D cylindrical model for precise vessel segmentation and radius estimation. This advanced method improves accuracy in medical imaging analysis compared to existing techniques.
Area of Science:
- Medical Imaging
- Image Processing
- Computational Anatomy
Background:
- Accurate 3-D segmentation and quantification of vessels are crucial for diagnosing and monitoring various medical conditions.
- Existing methods often struggle with precise vessel radius estimation and shape characterization in complex 3-D datasets.
Purpose of the Study:
- To introduce and validate a new 3-D cylindrical parametric intensity model for vessel segmentation and quantification.
- To evaluate the performance of the proposed model against established methods using synthetic and real-world medical imaging data.
Main Methods:
- Development of a 3-D cylindrical parametric intensity model.
- Direct fitting of the model to image intensities using an incremental Kalman filter-based process.
- Estimation of vessel centerline, radius, 3-D position, orientation, contrast, and fitting error.
- Validation using 3-D synthetic images and comparison with Gaussian model and randomized Hough transform approaches.
Main Results:
- The new approach accurately segments vessel centerlines and estimates local vessel radius, 3-D position, orientation, and contrast.
- Extensive validation on synthetic data demonstrated superior performance in vessel radius estimation compared to Gaussian and Hough transform methods.
- Successful application to 3-D Magnetic Resonance Angiography (MRA) and computed tomography angiography (CTA) image data, confirmed by radiologist-provided ground truth.
Conclusions:
- The proposed 3-D cylindrical parametric intensity model offers a robust and accurate method for 3-D vessel segmentation and quantification.
- This approach demonstrates significant improvements, particularly in estimating vessel radius, outperforming previous Gaussian and Hough transform-based techniques.
- The model's successful application to MRA and CTA data highlights its clinical relevance and potential for enhanced medical image analysis.
