Automated Brain MR Image Extraction Method
Author Information
Author(s): Zhang Haiyan, Liu Jiafeng, Zhu Zixin, Li Haiyun
Primary Institution: Capital Medical University
Hypothesis
Can an improved geometric active contour model provide an efficient and accurate method for brain extraction from MR images?
Conclusion
The proposed method can provide automated and accurate brain extraction results with high efficiency.
Supporting Evidence
- The method was compared to manual segmentation results and two popular methods, showing improved efficiency.
- It effectively addresses boundary leakage and intensity inhomogeneity issues in MR images.
- The algorithm was implemented on a standard PC and tested on 10 normal T1 MR brain images.
Takeaway
This study created a computer program that can automatically find and separate the brain from other parts in MRI pictures, making it easier and faster to analyze brain images.
Methodology
The method uses an improved geometric active contour model to extract brain regions from MR images, applying a binary level set function and local region-based signed pressure force.
Limitations
The method was only tested on normal adult MRI brain data sets, and further testing is needed on larger and more diverse samples.
Participant Demographics
Normal adult MRI brain data sets were used.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website