Marching Windows: Scalable Mesh Generation for Volumetric Data With Multiple Materials.

Area of Science:

• Computational geometry
• Medical imaging analysis
• Data visualization

Background:

• Volumetric datasets are rapidly growing in size due to improved imaging technologies.
• Existing tools struggle with large datasets, requiring excessive memory and often failing to preserve multi-material structures.
• Compact tetrahedral meshes are needed for efficient visualization and simulation of volumetric data.

Purpose of the Study:

• To develop a scalable method for computing high-quality tetrahedral meshes from large volumetric datasets.
• To address the memory and topological preservation limitations of current volumetric meshing techniques.
• To enable efficient processing and analysis of complex multi-material volumetric data.

Main Methods:

• Proposed a novel approach named Marching Windows.
• Implemented a moving window and disk-swap strategy to minimize runtime memory usage.
• Developed a new scheme to guarantee preservation of the original dataset's topological structure.
• Utilized an error-guided optimization technique for improved geometric approximation and mesh quality.

Main Results:

• The Marching Windows method successfully processed volumetric datasets exceeding the capacity of existing methods.
• Generated high-quality tetrahedral meshes that accurately represent multi-material structures.
• Demonstrated significant reduction in runtime memory footprint compared to traditional approaches.

Conclusions:

• Marching Windows offers a scalable and memory-efficient solution for meshing large volumetric datasets.
• The method preserves topological integrity and improves geometric accuracy, making it suitable for medical imaging and other fields.
• This approach enables advanced visualization and simulation of complex volumetric data previously intractable.