Related Experiment Video
Updated: Nov 7, 2025
05:41
Author Spotlight: Integrating Ultrasound Imaging with Biochemical Markers for Thyroid Disease Diagnosis
Published on: February 9, 2024
832
Local-global active contour model based on tensor-based representation for 3D ultrasound vessel segmentation
Jiahui Dong1, Danni Ai1, Jingfan Fan1
1Laboratory of Beijing Engineering Research Center of Mixed Reality and Advanced Display, School of Optics and Photonics, Beijing Institute of Technology, Beijing 100081, People's Republic of China.
Physics in Medicine and Biology
|April 28, 2021
Summary
This study introduces a novel 3D tensor-based active contour model for precise 3D vessel segmentation in ultrasound images. The method enhances accuracy, even with low-contrast and noisy data, improving visualization for physicians.
Area of Science:
- Medical Image Analysis
- Biomedical Engineering
Background:
- Accurate three-dimensional (3D) vessel segmentation is crucial for understanding vascular structures in medical imaging.
- Existing methods struggle with noise and low-contrast in 3D ultrasound (US) images.
Purpose of the Study:
- To develop an improved 3D vessel segmentation method for 3D ultrasound (US) images.
- Enhance segmentation accuracy and robustness in challenging imaging conditions.
Main Methods:
- A 3D tensor-based active contour model is proposed.
- The method incorporates contrast-independent multiscale bottom-hat tensor representation and local-global information.
- This approach addresses noise and low-contrast issues inherent in 3D US images.
Main Results:
- The proposed method achieved superior segmentation accuracy compared to state-of-the-art techniques on clinical 3D US and public 3D Multiphoton Microscopy datasets.
- Quantitative results showed mean values for SE, SP, and ACC of 0.7768 ± 0.0597, 0.9978 ± 0.0013, and 0.9971 ± 0.0015, respectively.
- The method successfully segmented complex vessels in noisy, low-contrast medical images.
Conclusions:
- The developed 3D tensor-based active contour model provides smoother and more accurate vessel boundaries.
- This method offers a robust solution for 3D vessel segmentation in challenging medical imaging scenarios.
- The approach has significant potential for various medical image processing and analysis applications.
