Level Set Based Hippocampus Segmentation in MR Images with Improved Initialization Using Region Growing

Area of Science:

• Neuroimaging
• Medical Image Analysis
• Computational Anatomy

Background:

• The hippocampus is vital in neurological disorders like Alzheimer's disease.
• Accurate hippocampus segmentation in MR images is challenging due to anatomical complexities and image quality.
• Existing segmentation methods often struggle with the hippocampus's intricate shape and boundaries.

Purpose of the Study:

• To develop an accurate and efficient method for segmenting the hippocampus from MR images.
• To improve the detection of neurological disorders through precise hippocampal volume analysis.

Main Methods:

• A novel image segmentation approach integrating adaptive region growing with a level set algorithm.
• Utilizing morphological operations to refine initial contours for level set evolution.
• Employing an improved edge-based level set method with global Gaussian distributions for enhanced accuracy.
• Applying gradient descent for energy equation minimization.

Main Results:

• The proposed method successfully segmented the hippocampus with high accuracy.
• Segmentation contours closely approximated manual segmentations performed by specialists.
• The technique demonstrated efficiency in handling challenging segmentation scenarios.

Conclusions:

• The combined adaptive region growing and level set method offers a robust solution for hippocampus segmentation.
• This approach holds promise for improving the diagnosis and monitoring of Alzheimer's disease and related conditions.
• The method's accuracy supports its potential clinical application in neurological assessments.