Breast Window Definition Using Edge Search and Thresholding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Radiologists face inefficiencies and delays when manually adjusting mammography images to define a breast window, as current methods rely on computationally intensive edge detection techniques, impacting the speed and accuracy of image analysis.
Innovation Solution
A computer-implemented method and system that identifies a center region of the breast image, defines image edges as borders, and searches along these edges to define a breast window, using subsampling and thresholding to efficiently identify breast boundaries without traditional edge detection, thereby enhancing responsiveness and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional edge detection techniques are used to identify the breast boundary, then the breast window definition is accurate, but the computational complexity increases and processing speed decreases
Solution Approach 1:
The patent segments the breast boundary identification process into distinct phases: initial center region identification, directional searching along image edges, and iterative refinement. This segmentation allows the system to avoid computationally intensive global edge detection while maintaining accurate boundary identification through localized, sequential processing steps.
Solution Approach 2:
The patent performs preliminary actions by first identifying the center region of the breast image and establishing initial search directions along image edges before conducting the actual boundary searching. This preliminary setup enables the system to constrain subsequent boundary detection to specific regions and directions, significantly reducing the computational search space while maintaining accuracy.
2Measurement precision
If traditional edge detection techniques are used to identify the breast boundary, then the breast window definition is accurate, but the processing time increases
Solution Approach 1:
The patent segments the breast boundary identification process into distinct phases: initial center region identification, directional searching along image edges, and iterative refinement. This segmentation allows the system to avoid computationally intensive global edge detection while maintaining accurate boundary identification through localized, sequential processing steps.
Solution Approach 2:
The patent performs preliminary actions by first identifying the center region of the breast image and establishing initial search directions along image edges before conducting the actual boundary searching. This preliminary setup enables the system to constrain subsequent boundary detection to specific regions and directions, significantly reducing the computational search space while maintaining accuracy.
3Area of stationary object
If manual adjustment of zoom factor is used to fill the viewing area with the breast image, then the viewing area is properly configured, but the operation becomes inconvenient and time-consuming
Solution Approach 1:
The patent implements self-service by enabling the system to automatically perform the entire breast window definition process without requiring manual radiologist intervention. The system autonomously identifies the breast boundary, calculates the optimal viewing area, and configures the display parameters, thereby eliminating the need for manual zoom factor adjustment and significantly improving operational convenience while maintaining proper viewing area configuration.
Solution Approach 2:
The patent utilizes parameter changes by dynamically adjusting display parameters such as zoom factor, pan position, and windowing based on automatically detected breast boundary characteristics. This allows the system to transform the static manual adjustment process into a dynamic automated process that adapts parameters according to the specific breast image geometry, achieving both proper viewing area configuration and operational efficiency.
Data Source
AI summary
A computer implemented method for defining a breast window within an image for a breast, said image containing a breast image having a breast boundary, the method comprising: identifying a center region of the breast image, where the center region is coincident with an edge of the image; defining the edge of the image as the first border of the breast window; searching from the center region along the edge of the image for the breast boundary, the breast boundary informing a second border of the breast window; searching from the center region along the edge of the image in an opposite direction for the breast boundary, the breast boundary informing a third border of the breast window; and repeatedly searching from the center region in a direction orthogonal to the edge of the image for a point on the breast boundary most distant from the edge of the image, the point on the breast boundary informing a fourth border of the breast window. In some embodiments, the image may be rotated/flipped and/or subsampled and/or thresholded to form a binary image before a breast window is searched. A method of thresholding the image is also disclosed.