Related Experiment Video
Updated: May 30, 2026
06:48
Automated Segmentation of Cortical Grey Matter from T1-Weighted MRI Images
Published on: January 7, 2019
Manual refinement system for graph-based segmentation results in the medical domain
Jan Egger1, Rivka R Colen, Bernd Freisleben
1Department of Radiology, Harvard Medical School, Boston, MA 02115, USA. egger@bwh.harvard.edu
Journal of Medical Systems
|August 10, 2011
Summary
This study introduces a manual refinement technique for graph-based image segmentation. By incorporating user-defined seed points, it enhances segmentation accuracy and efficiency in medical imaging.
Area of Science:
- Computer Vision
- Medical Image Analysis
- Computational Imaging
Background:
- Graph-based methods segment images by modeling them as graphs, using pixel intensity differences for edge weights.
- Image segmentation, particularly in medical contexts, often requires manual refinement due to inherent limitations of automated methods.
Purpose of the Study:
- To develop an intuitive and efficient manual refinement method for graph-based image segmentation.
- To improve the accuracy and user-friendliness of segmentation results, especially for medical images.
Main Methods:
- The proposed method refines graph-cut segmentation by incorporating user-defined seed points as fixed nodes within the graph.
- This approach leverages the existing structure of graph-based segmentation algorithms.
Main Results:
- Demonstrated feasibility of the manual refinement approach on both synthetic and real image datasets.
- The method allows for intuitive and rapid integration of manual edits into automated segmentation.
Conclusions:
- The presented manual refinement technique offers a practical enhancement to graph-based image segmentation.
- It effectively addresses the need for post-processing adjustments in sensitive applications like medical image segmentation.
