Related Experiment Video
Updated: May 11, 2026
11:03
High-resolution In Vivo Manual Segmentation Protocol for Human Hippocampal Subfields Using 3T Magnetic Resonance Imaging
Published on: November 10, 2015
Fully-automated approach to hippocampus segmentation using a graph-cuts algorithm combined with atlas-based
Kichang Kwak1, Uicheul Yoon, Dong-Kyun Lee
1Department of Biomedical Engineering, Hanyang University, Seoul, South Korea.
Magnetic Resonance Imaging
|May 21, 2013
Summary
This study introduces an automated method for precise hippocampus segmentation, crucial for diagnosing Alzheimer's disease (AD). The novel approach significantly improves accuracy and reliability compared to traditional techniques.
Area of Science:
- Neuroimaging
- Medical Image Analysis
- Computational Neuroscience
Background:
- The hippocampus is a key biomarker for neurological and psychiatric diseases like Alzheimer's disease (AD).
- Accurate, robust, and reproducible segmentation of hippocampal structures is essential for clinical applications.
- Existing segmentation methods may lack the required precision for reliable biomarker analysis.
Purpose of the Study:
- To develop and validate an automated hippocampal segmentation method.
- To enhance the accuracy, robustness, and reproducibility of hippocampus delineation.
- To compare the proposed method against conventional atlas-based segmentation.
Main Methods:
- Proposed an automated segmentation method combining graph-cuts, atlas-based segmentation, and morphological opening.
- Utilized atlas-based segmentation for initial region definition and a priori information.
- Incorporated partial volume probability estimation for seed definition and morphological opening to reduce false positives.
Main Results:
- The proposed method achieved a higher similarity index (0.81±0.03) compared to conventional atlas-based segmentation (0.72±0.04).
- Demonstrated improved segmentation accuracy with higher precision (0.76±0.04) and recall (0.86±0.05) than conventional methods.
- Experiments on twenty-seven healthy subjects confirmed the method's reliability and plausibility.
Conclusions:
- The developed automated method provides accurate, robust, and reliable hippocampus segmentation.
- This technique offers a significant advancement over traditional methods for clinical and research applications.
- The improved segmentation accuracy supports its potential use in diagnosing and monitoring diseases like Alzheimer's.
