Related Experiment Video
Updated: Jan 16, 2026
A Full Skin Defect Model to Evaluate Vascularization of Biomaterials In Vivo
Published on: August 28, 2014
Optimized Vessel Segmentation: A Structure-Agnostic Approach With Small Vessel Enhancement and Morphological
Insights
OVS-Net offers improved blood vessel segmentation by enhancing small vessel identification and preserving connectivity. This optimized framework generalizes across diverse imaging, outperforming existing methods for clinical applications.
Area of Science:
- Medical Imaging
- Computer Vision
- Biomedical Engineering
Background:
- Accurate blood vessel segmentation is crucial for clinical analysis but faces challenges like sparsity, low contrast, and maintaining topological integrity.
- Existing specialized and general-purpose segmentation models struggle with generalization and preserving vascular connectivity.
Purpose of the Study:
- To introduce OVS-Net, an optimized vessel segmentation framework designed for improved generalization across diverse vascular structures and imaging modalities.
- To address limitations in small vessel segmentation and ensure the preservation of vascular topology and connectivity.
Main Methods:
- Developed OVS-Net featuring a dual-branch architecture for enhanced small vessel segmentation.
- Incorporated a morphology-aware correction module to maintain vascular topology and connectivity.
- Compiled a multi-modality dataset from 17 sources for comprehensive training and benchmarking.
Main Results:
- OVS-Net demonstrated superior segmentation accuracy and generalization capabilities compared to 6 SAM-based and 17 expert models.
- Achieved a significant 34.6% improvement in segmentation connectivity.
- Validated performance across various imaging conditions and vessel types.
Conclusions:
- OVS-Net provides a robust and generalizable solution for blood vessel segmentation.
- The framework's ability to preserve connectivity and handle diverse vascular structures shows significant potential for clinical applications.
- The availability of code and dataset facilitates further research and development.
Abstract:
Accurate segmentation of blood vessels is essential for various clinical assessments and postoperative analyses. However, the inherent challenges of vascular imaging-such as sparsity, fine granularity, low contrast, data distribution variability, and the critical need for preserving topological integrity-make generalized vessel segmentation particularly complex. While specialized segmentation methods have been developed for specific anatomical regions, their over-reliance on tailored models hinders broader applicability and generalization. General-purpose segmentation models introduced in medical imaging often fail to address critical vascular characteristics, including the connectivity of segmentation results. In this study, we propose OVS-Net, an optimized vessel segmentation framework designed to generalize across diverse vessel structures and imaging modalities. It introduces a dual-branch architecture design for improving small vessel segmentation and a morphology-aware correction module to preserve vascular topology and connectivity. We compiled a comprehensive multi-modality dataset from 17 datasets to train and benchmark the proposed OVS-Net against 6 SAM-based methods and 17 expert models under various conditions. The results demonstrate that our approach achieves superior segmentation accuracy, generalization, and a 34.6% improvement in connectivity, underscoring its potential for clinical applications. The code and dataset information are available at https://github.com/Hk416mod2/OVS-Net.
