Breast Tomosynthesis Lesion Diagnosis Using Priority Target Regions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for diagnosing breast lesions using radiographic images, such as those described in WO2014/192187A, face challenges in accurately determining whether structures are lesions, particularly when mammary gland structures overlap.
Innovation Solution
An image processing device and method that utilizes tomosynthesis imaging to detect specific structural patterns in breast images, synthesizes multiple tomographic images into a two-dimensional image, specifies a priority target region, and performs lesion diagnosis based on this synthesized image, using machine learning models to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional radiographic image analysis is used for lesion determination, then the diagnostic process is simple, but the accuracy of determining whether structures are lesions is insufficient
Solution Approach 1:
The patent segments the breast tissue into multiple tomographic images at different depths, allowing lesion determination to be performed on specific layers rather than analyzing the entire overlapping structure in a single radiographic image. This segmentation enables more accurate lesion identification by isolating structures in three-dimensional space.
Solution Approach 2:
The patent transitions from two-dimensional radiographic imaging to three-dimensional tomosynthesis by acquiring images at multiple angles and reconstructing tomographic slices. This dimensional change allows differentiation of overlapping structures along the depth axis, significantly improving lesion determination accuracy.
2Measurement precision
If tomosynthesis imaging is used to obtain multiple tomographic images, then lesion determination accuracy is improved, but the processing time and computational load increase
Solution Approach 1:
The patent performs preliminary processing by pre-segmenting the tomosynthesis data into multiple tomographic images and pre-identifying candidate regions in each layer before the final lesion determination. This preliminary action reduces the computational burden during the actual diagnosis phase, decreasing processing time while maintaining accuracy.
Solution Approach 2:
The patent extracts and focuses analysis on specific priority target regions where lesions are most likely to be present, rather than processing the entire volume of tomographic data. This extraction of relevant regions reduces computational load and processing time while preserving diagnostic accuracy.
3Measurement precision
If the entire synthesized two-dimensional image is analyzed for lesion determination, then comprehensive coverage is achieved, but the determination accuracy for specific lesions is reduced due to overlapping structures
Solution Approach 1:
The patent applies different analysis strategies to different regions of the breast by identifying priority target regions with high lesion probability. Instead of uniform analysis across the entire image, the system focuses computational resources on specific local areas where lesions are most likely to occur, improving determination accuracy for those critical regions.
Solution Approach 2:
The patent performs detailed lesion determination analysis only on priority target regions rather than the entire synthesized image. This partial action concentrates diagnostic effort on the most suspicious areas, achieving high accuracy for lesion determination while reducing the effective analysis area and processing requirements.
Data Source
AI summary
An image processing device includes at least one processor. The processor detects a specific structural pattern indicating a lesion candidate structure for a breast in a series of a plurality of projection images obtained by performing tomosynthesis imaging on the breast or in a plurality of tomographic images obtained from the plurality of projection images, synthesizes the plurality of tomographic images to generate a synthesized two-dimensional image, specifies a priority target region, in which the specific structural pattern is present, in the synthesized two-dimensional image, and performs determination regarding a diagnosis of a lesion on the basis of the synthesized two-dimensional image and the priority target region.