Segmenting Bullous and Diffuse Emphysema in CT Volumes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current imaging techniques, particularly in computed tomography, struggle to effectively visualize irregularly shaped features of emphysema subtypes like bullous emphysema, which can be occluded and overlooked in standard renderings based on voxel intensities.
Innovation Solution
An image data processor is used to identify voxels corresponding to bullous and diffuse emphysema, generating 2D projection images where the intensity of bullae contours is based on size, and highlighting these subtypes with different indicia, such as colors, to create a composite image that visually distinguishes both subtypes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If standard minimum intensity projection or mean projection rendering is used based on voxel intensities, then diffuse emphysema can be visualized, but bullous emphysema features are occluded and overlooked
Solution Approach 1:
The patent segments the emphysema visualization into two separate 2D projection images: one for diffuse emphysema and one for bullous emphysema. Each image is generated by processing voxels with different criteria (intensity-based for diffuse, size-based for bullous), allowing each subtype to be visualized without occlusion from the other.
Solution Approach 2:
The patent applies different processing qualities to different regions: diffuse emphysema is highlighted using intensity-based voxel processing, while bullous emphysema is highlighted using size-based contour processing. This local differentiation allows each subtype to be optimized for its specific visual characteristics.
2Ease of manufacture
If single 2D projection image is generated using standard voxel intensity thresholding, then processing is simple, but irregularly shaped bullous features cannot be visually distinguished
Solution Approach 1:
The patent divides the single 2D projection into two separate images processed with different algorithms. The diffuse emphysema image uses standard intensity thresholding, while the bullous emphysema image uses size-based contour detection, enabling precise visual distinction of each subtype.
Solution Approach 2:
The patent changes the projection parameters based on the emphysema subtype: for diffuse emphysema, intensity thresholding is used; for bullous emphysema, size-based contour parameters are used. This parameter adaptation enables precise visualization of each subtype's characteristic features.
3Ease of operation
If only diffuse emphysema is visualized using standard rendering, then the rendering process is straightforward, but bullous emphysema subtypes are occluded
Solution Approach 1:
The patent segments the visualization process into two independent 2D projection image generation steps, each optimized for its target subtype. This segmentation prevents information loss by ensuring both diffuse and bullous emphysema are captured in their respective images.
Solution Approach 2:
The patent merges the two separate 2D projection images (diffuse and bullous emphysema) into a single composite 2D projection image that displays both subtypes simultaneously. This merging preserves all emphysema information while maintaining the ease of viewing a single integrated image.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An processor (118) includes a bullous emphysema identifier (206) that processes voxels of the volumetric image data and identifies voxels corresponding to bullous emphysema, a two dimensional projection image generator (206) that generates a 2D bullous emphysema projection image based on the voxels corresponding to bullous emphysema, wherein an intensity of a contour of a bulla in the 2D bullous emphysema projection image is based on a size of the bulla, and a feature highlighter (210) that highlights the bullous emphysema in the 2D bullous emphysema projection image using second first indicia.