DECODE-3DViz: Efficient WebGL-Based High-Fidelity Visualization of Large-Scale Images using Level of Detail and Data...

Area of Science:

• Medical Imaging Visualization
• Computer Graphics
• Web Technologies

Background:

• Rendering high-resolution volumetric medical data on the web presents significant challenges.
• Existing WebGL solutions often struggle with large datasets, browser memory, and texture size limitations.
• Efficient visualization is crucial for accurate diagnosis in medical imaging, particularly for complex structures like the peripheral vasculature.

Purpose of the Study:

• To introduce the DECODE-3DViz pipeline for advanced web-based visualization of large-scale medical imaging data.
• To address the technical hurdles in rendering high-resolution volumetric datasets using WebGL.
• To optimize real-time interaction and high-fidelity visualization for peripheral artery computed tomography (CT) images.

Main Methods:

• Integration of progressive chunk streaming and level of detail (LOD) algorithms.
• Optimization for WebGL texture size constraints and browser memory limitations.
• Comparative performance evaluation against existing state-of-the-art visualization tools.

Main Results:

• Achieved up to a 98% reduction in rendering time compared to competitors.
• Maintained high frame rates up to 144 FPS.
• Demonstrated exceptional GPU memory efficiency, using as little as 2.6 MB, significantly less than the >100 MB of other tools.
• User feedback indicated high satisfaction (4.3/5) with performance, structure definition, and diagnostic capability.

Conclusions:

• DECODE-3DViz offers a significant advancement in web-based medical imaging visualization.
• The pipeline enables detailed and accurate visualization of peripheral vasculature, potentially improving diagnostic accuracy and clinical outcomes.
• The open-source nature of DECODE-3DViz promotes accessibility and further development in the field.