Segmented Image Display with Multi-Plane Boundaries
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manually reviewing numerous two-dimensional images to verify the full extent of a tumor for accurate reporting and analysis is time-consuming and prone to errors, especially for tumors with ill-defined shapes extending beyond single image planes.
Innovation Solution
A method and system that automatically display a reference two-dimensional image with segmentation boundaries derived from both the current and non-reference image planes, allowing for concurrent visualization of segmentation boundaries across multiple planes, reducing the need for manual review of multiple images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If numerous two-dimensional images are manually reviewed to verify tumor extent, then diagnostic accuracy is improved, but time consumption and operational difficulty increase
Solution Approach 1:
The patent merges segmentation boundaries from multiple image planes into a single reference image display. The system combines first segmentation boundaries from the reference image plane with second segmentation boundaries from non-reference image planes, allowing users to view tumor extent across multiple planes simultaneously in one image rather than reviewing numerous separate images.
Solution Approach 2:
The patent transitions from a two-dimensional sequential review approach to a multi-dimensional integrated display. By overlaying segmentation boundaries from different image planes (sagittal, axial, coronal) onto the reference image, the system adds spatial dimensionality to the display, enabling users to perceive tumor extent in three-dimensional space through a single two-dimensional image.
2Measurement precision
If numerous two-dimensional images are manually reviewed to verify tumor extent, then diagnostic accuracy is improved, but ease of operation deteriorates
Solution Approach 1:
The patent merges segmentation boundaries from multiple image planes into a single reference image display. The system combines first segmentation boundaries from the reference image plane with second segmentation boundaries from non-reference image planes, allowing users to view tumor extent across multiple planes simultaneously in one image rather than reviewing numerous separate images.
Solution Approach 2:
The system automatically generates and displays the integrated segmentation boundaries without requiring manual intervention to review multiple images. The computer automatically performs the segmentation boundary extraction and combination process, eliminating the need for users to manually navigate and compare numerous images, thereby significantly improving operational ease.
3Ease of operation
If segmentation boundaries from multiple planes are integrated into a single image, then operational ease is improved, but complexity of the display system increases
Solution Approach 1:
The patent applies segmentation by dividing the tumor boundary into distinct components: first segmentation boundaries from the reference image plane and second segmentation boundaries from non-reference image planes. This segmentation allows the system to process and display different spatial information separately before integrating them, managing complexity through structured decomposition of the segmentation task.
Solution Approach 2:
The reference image serves as an intermediary载体 that holds both the original image data and the overlaid segmentation boundaries from multiple planes. This intermediary approach allows the system to manage complexity by using the reference image as a stable foundation upon which multiple segmentation datasets are layered, simplifying the integration process while maintaining operational ease.
Data Source
AI summary
A method for displaying a segmented two-dimensional (2D) image includes obtaining a three-dimensional (3D) volume dataset corresponding to an imaged volume along a viewing plane, segmenting an object of interest within 3D volume to generate a plurality of segmented two-dimensional (2D) images along the viewing plane, selecting a reference image for viewing from the plurality of segmented 2D images, and displaying the reference image, the reference image having a first segmentation boundary drawn around the object of interest and a second segmentation boundary drawn around the object of interest, the first segmentation boundary being derived from the segmentation performed on the reference image and the second segmentation boundary being derived from the segmentation performed on at least one non-reference image of the plurality of segmented 2D images.