Model-Based Segmentation Correcting Partial Volume Effects
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Medical imaging techniques such as MRI, CT, and PET face challenges in accurate segmentation due to partial volume effects, where a single voxel contains tissue from multiple anatomical regions, leading to inaccurate image classification.
Innovation Solution
A model-based segmentation method using a surface mesh is employed, where the mesh is deformed to fit the medical image, and the position of its faces is adjusted along outward-facing vectors to correct voxel intensities, ensuring they reflect the fractions of different tissue types, thereby addressing partial volume effects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If model-based segmentation is used to divide medical images into discrete volumes, then segmentation can be performed, but partial volume effects occur at boundaries causing inaccurate classification
Solution Approach 1:
The patent applies partial volume correction by calculating the proportion of each tissue type within boundary voxels rather than making binary classification. The surface mesh is positioned to intersect voxels at fractional positions, and the intensity is corrected based on the fraction of each tissue type present, rather than requiring complete separation of tissues.
Solution Approach 2:
The patent changes the parameter representation from discrete binary classification to continuous fractional representation. Boundary voxels are assigned intensity values that reflect the fractional composition of different tissue types, transforming the segmentation from a discrete problem to a continuous one that better represents physical reality.
2Measurement precision
If surface mesh is deformed to fit anatomical regions, then segmentation accuracy improves, but partial volume effects at boundaries still cause intensity inaccuracies
Solution Approach 1:
The patent uses an iterative optimization process where the surface mesh position is adjusted based on feedback from the partial volume correction calculation. The mesh position and tissue fraction estimates are refined mutually until convergence, allowing the segmentation to adapt to the actual tissue distribution while correcting for partial volume effects.
Solution Approach 2:
The patent introduces tissue fraction estimates as an intermediary variable that mediates between the surface mesh position and the voxel intensity values. This intermediary allows the system to account for partial volume effects without requiring the mesh to be positioned exactly at tissue boundaries, resolving the conflict between anatomical accuracy and intensity accuracy.
Data Source
AI summary
The invention provides for a medical apparatus (100, 400, 600) comprising a memory (110) for storing machine executable instructions (120) and a processor (104) for controlling the medical apparatus. Execution of the machine executable instructions causes the processor to: receive (200) a medical image (122) descriptive of a three-dimensional anatomy of a subject (418); and provide (202) an image segmentation (124) by segmenting the medical image into multiple tissue regions (300, 302) using a model-based segmentation. The model-based segmentation assigns a tissue type to each of the multiple regions. The model-based segmentation has a surface mesh (304). The segmentation is corrected by using the tissue type assigned to each of the multiple regions to correct for partial volume effects at boundaries formed by the surface mesh between at least some of the multiple tissue regions.