3D Live-Wires on mosaic volumes.

Area of Science:

• Medical Imaging
• Computer Vision
• Image Segmentation

Background:

• Accurate segmentation of anatomical structures in 3D medical images is crucial for diagnosis and treatment planning.
• Existing segmentation methods often require extensive parameter tuning or problem-specific knowledge.
• The 3D Live-Wire technique offers a novel approach for interactive segmentation.

Purpose of the Study:

• To evaluate the performance of the 3D Live-Wire technique for segmenting over-segmented 3D mosaic volumes.
• To assess the accuracy and efficiency of the method in segmenting white cerebral matter from MRI data.
• To determine the robustness and intuitiveness of parameter selection across different tissues and modalities.

Main Methods:

• The 3D Live-Wire technique was applied to pre-processed 3D mosaic volumes using histogram spreading and region growing for over-segmentation.
• Interactive segmentation was performed using user-defined seed points, generating a triangle mesh via Delaunay triangulation.
• Closed surface patches were created for each triangle using the 3D Live-Wire algorithm.

Main Results:

• The technique achieved a mean deviation of approximately 0.60 pixels from the correct boundary for white cerebral matter segmentation.
• The average segmentation time was recorded at 229.5 seconds.
• The method demonstrated robust and intuitive parameter selection without requiring specialized knowledge.

Conclusions:

• The 3D Live-Wire technique provides an effective and efficient solution for interactive 3D image segmentation.
• Its ability to handle over-segmented volumes and intuitive parameter selection makes it suitable for various medical imaging applications.
• Further validation across diverse datasets and anatomical structures is warranted.