Related Experiment Video
Updated: May 28, 2026
09:06
Whole-brain Segmentation and Change-point Analysis of Anatomical Brain MRI—Application in Premanifest Huntington's Disease
Published on: June 9, 2018
Segmenting images by combining selected atlases on manifold.
Yihui Cao1, Yuan Yuan, Xuelong Li
1Center for OPTical IMagery Analysis and Learning (OPTIMAL), State Key Laboratory of Transient Optics and Photonics, Xi'an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi'an 710119, Shaanxi, P.R. China.
Medical Image Computing and Computer-Assisted Intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention
|October 19, 2011
Summary
This study introduces a novel manifold projection method for improved medical image segmentation. It enhances atlas selection and combination, leading to more accurate segmentation results, particularly with large datasets.
Area of Science:
- Medical image analysis
- Computer vision
- Computational anatomy
Background:
- Atlas selection and combination are crucial for atlas-based segmentation performance.
- Current methods in original image space may not accurately capture intrinsic image similarity.
- This can lead to suboptimal atlas selection and misleading template generation for segmentation.
Purpose of the Study:
- To propose a novel method for atlas selection and combination using low-dimensional manifold projection.
- To improve the accuracy and robustness of atlas-based medical image segmentation.
- To develop an efficient weighting method for combining selected atlases.
Main Methods:
- Projecting images onto a low-dimensional manifold to assess intrinsic similarity.
- Selecting atlases based on their proximity in the low-dimensional space.
- Implementing a novel weighting scheme for atlas combination.
Main Results:
- The proposed manifold projection approach enhances atlas selection accuracy.
- The novel weighting method improves the efficiency of atlas combination.
- Experimental results show robust and accurate segmentation performance, especially with large training sets.
Conclusions:
- Manifold projection offers a superior approach for atlas selection and combination in segmentation.
- The method improves segmentation accuracy and robustness.
- This technique is particularly beneficial for large-scale medical image analysis tasks.
