Volumetric Dataset Rendering via Segmented Data Sizes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Rendering high-resolution volumetric datasets from medical imaging, such as CT scans, is challenging due to memory limitations in host or GPU memory, especially when using computationally intensive methods like physically based rendering.
Innovation Solution
A computer-implemented method that involves obtaining two types of volume data representing different portions of a volumetric dataset, where the second data has a lower data size per unit volume than the first. A physically based rendering process, such as path tracing, is performed on a combination of these data to generate visual parameter data for rendering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If physically based rendering processes are used to render high-resolution volumetric datasets, then rendering quality and realism are improved, but memory requirements increase beyond host or GPU memory limitations
Solution Approach 1:
The volumetric dataset is divided into multiple portions, each represented by volume data at different data sizes per unit volume. The rendering process segments the rendering task across these portions, using higher data size portions for critical regions and lower data size portions for less critical regions, thereby reducing overall memory requirements while maintaining rendering quality.
Solution Approach 2:
Different portions of the volumetric dataset are represented by volume data with different data sizes per unit volume. The system selectively uses higher data size volume data where it is most needed for high-quality rendering and lower data size volume data where rendering quality can be compromised, optimizing the balance between rendering quality and memory usage.
2Measurement precision
If high-resolution volumetric datasets are loaded into memory for rendering, then rendering accuracy is improved, but host or GPU memory limitations are exceeded
Solution Approach 1:
The volumetric dataset is segmented into multiple portions with corresponding volume data at different resolutions. The system loads and processes only the necessary portions into memory rather than loading the entire high-resolution dataset, reducing memory capacity requirements while maintaining rendering accuracy for the processed regions.
Solution Approach 2:
Instead of loading the complete high-resolution volumetric dataset into memory, the system uses partial volume data representations that provide sufficient accuracy for rendering purposes. This partial action approach reduces memory requirements while achieving adequate rendering accuracy without the excessive memory consumption of full-resolution loading.
Data Source
AI summary
A computer-implemented method for use in rendering a volumetric dataset representing a volume, comprises: obtaining first volume data representing a first portion of the volume represented by the volumetric dataset; obtaining second volume data representing a second portion of the volume represented by the volumetric dataset, wherein at least some of the second portion of the volume is outside of the first portion of the volume, and wherein a second data size per unit volume of the second volume data is lower than a first data size per unit volume of the first volume data; and performing a first physically based rendering process on a combination of the first volume data and the second volume data to generate first visual parameter data for rendering the volumetric dataset.