3D Image Feature Enhancement via Multi-View Projection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Medical practitioners face challenges in precisely identifying anatomical features in 3D images due to irrelevant information hiding relevant features, which complicates diagnostic and intervention planning in medical imaging.
Innovation Solution
A method and system that enhance the display of features of interest in 3D images by obtaining and projecting 3D images from multiple viewpoints, defining regions of interest, and filtering voxels to remove irrelevant information, allowing for improved visualization of anatomical structures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If a 3D image is projected to display anatomical region information, then comprehensive anatomical information is provided, but irrelevant information hides relevant features making identification difficult
Solution Approach 1:
The patent segments the 3D volumetric data by creating multiple 2D planar sections (axial, sagittal, coronal views) from the 3D image. Each 2D image represents a specific slice through the anatomical region, allowing practitioners to examine specific planes without being obscured by overlapping structures in other planes. This segmentation transforms the complex 3D visualization problem into manageable 2D sections that can be displayed and analyzed separately.
Solution Approach 2:
The patent transitions from 3D volumetric display to 2D planar projections by projecting the 3D image data onto 2D planes at different orientations (axial, sagittal, coronal). This dimensionality reduction eliminates the occlusion problem inherent in 3D views by displaying anatomical structures in separate 2D slices, where structures that would overlap in 3D are separated in the 2D projection planes.
2Measurement precision
If multiple 2D images are projected from different viewpoints, then relevant features become visible, but processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary processing by pre-generating the complete set of 2D planar projections (axial, sagittal, coronal) from the 3D volumetric data before the actual diagnostic examination. These pre-computed 2D images are stored and can be rapidly displayed during clinical use without requiring real-time reconstruction, thus reducing processing time during the actual diagnostic workflow while maintaining high measurement precision.
Data Source
Figure 1
Figure 2
Figure 3a~3b
AI summary
The invention concerns a computer implemented method for enhancing the display of features of interest in a 3D image of an anatomical region of a patient, the method comprising: a) obtaining a 3D image of an anatomical region of a patient; b) projecting the 3D image according to a first point of view for obtaining a first 2D image; b1) projecting the 3D image according to a second point of view for obtaining a second 2D image of said anatomical region; the first point of view being transverse to the second point of view and the first 2D image being transverse to the second 2D image; the method comprising c) determining from the first 2D image a first 3D region of interest; d) projecting said 3D region of interest according to the second point of view for obtaining a third 2D image.