Related Experiment Video
Updated: May 3, 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
11.0K
Automatic structural parcellation of mouse brain MRI using multi-atlas label fusion
Da Ma1, Manuel J Cardoso2, Marc Modat2
1Centre for Medical Imaging Computing, University College London, London, England, United Kingdom ; Centre for Advanced Biomedical Imaging, Division of Medicine, University College London, London, England, United Kingdom.
Plos One
|January 30, 2014
Summary
This study introduces an automated mouse brain MRI segmentation framework using multi-atlas methods. The new approach significantly improves segmentation accuracy compared to existing single-atlas and original STAPLE techniques.
Area of Science:
- Neuroimaging
- Medical Image Analysis
- Computational Biology
Background:
- Multi-atlas segmentation propagation is a leading technique for automatic image parcellation.
- Its application in preclinical research, particularly for mouse brain MRI, remains limited.
Purpose of the Study:
- To develop and validate a fully automatic framework for mouse brain MRI structural parcellation.
- To enhance segmentation accuracy using multi-atlas segmentation propagation.
Main Methods:
- The framework utilizes the Similarity and Truth Estimation for Propagated Segmentations (STEPS) algorithm.
- It employs a locally normalized cross-correlation metric for atlas selection and an extended Simultaneous Truth and Performance Level Estimation (STAPLE) for multi-label fusion.
- Optimized parameters were determined using publicly available mouse brain atlases.
Main Results:
- The multi-atlas framework demonstrated significantly higher segmentation accuracy than single-atlas methods.
- It also outperformed the original STAPLE framework in accuracy.
- Optimized STEPS parameters improved label fusion performance.
Conclusions:
- The developed multi-atlas framework provides a robust and accurate solution for automatic mouse brain MRI parcellation.
- This methodology advances preclinical neuroimaging analysis.
- The optimized approach offers superior performance over existing methods.
