3D Sparse Volume Segmentation for Medical Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
As 3D scanners produce thinner slices, the processing time for data increases, exceeding the capacity of even powerful computers, necessitating a method for efficient sparse volume segmentation.
Innovation Solution
A method for sparse volume segmentation using a 3D sparse model that learns prior knowledge, selects key contours, builds a 3D sparse model, and interpolates the volume data, focusing the workload on critical points while ensuring robustness through prior knowledge constraints.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the scanner resolution is improved and slice thickness is reduced, then the image quality and detail are improved, but the processing time increases and computational capacity is exceeded
Solution Approach 1:
The patent applies segmentation by dividing the volumetric data into a sparse set of key contours at selected indices rather than processing all volume data. This selective segmentation approach processes only critical cross-sections while inferring intermediate sections, directly resolving the contradiction by reducing processing time while maintaining measurement precision through strategic sampling of key anatomical levels
Solution Approach 2:
The patent implements partial action by processing only a subset of key contours rather than the complete volume data. By selecting representative key indices and contours that capture essential anatomical features, the system achieves sufficient segmentation accuracy with significantly reduced computational effort, balancing measurement precision requirements against processing time constraints
2Manufacturing precision
If the number of slices for a given organ increases, then the segmentation accuracy is improved, but the computational load exceeds computer capacity
Solution Approach 1:
The patent extracts only the essential key contours from the volumetric data at strategically selected indices, removing the need to process all intermediate slices. This extraction approach maintains segmentation accuracy by focusing on critical anatomical levels while dramatically reducing computational load to within available computer capacity
Solution Approach 2:
The patent performs preliminary action by pre-selecting key indices and contours that capture the most important anatomical information before full segmentation. This preliminary selection of representative cross-sections enables accurate segmentation inference without requiring computational processing of the complete high-resolution volume data
Data Source
AI summary
A computer readable medium is provided embodying instructions executable by a processor to perform a method for sparse volume segmentation for 3D scan of a target. The method including learning prior knowledge, providing volume data comprising the target, selecting a plurality of key contours of the image of the target, building a 3D spare model of the image of the target given the plurality of key contours, segmenting the image of the target given the 3D sparse model, and outputting a segmentation of the image of the target.