CAD Indicator Icons for Breast Tomosynthesis Slice Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge in breast cancer screening lies in effectively and efficiently reviewing large volumes of data from x-ray tomosynthesis images, where CAD markers can be easily overlooked due to the sheer number of slices, leading to potential missed detections despite advancements in CAD systems for conventional mammography.
Innovation Solution
An interactive user interface is developed to process and display CAD information in breast x-ray tomosynthesis data volumes, featuring graphical navigation tools, CAD indicator icons, and fibroglandular tissue density maps to prioritize and highlight relevant findings, ensuring consistent marking of CAD findings across cases and volumes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional x-ray mammography is used for breast cancer screening, then the screening process is simple and fast, but cancers may be missed due to obscuration by dense fibroglandular tissue
Solution Approach 1:
The patent transitions from two-dimensional projection mammography to three-dimensional tomosynthesis imaging. The x-ray source rotates through a limited angle (e.g., -15 to +15 degrees) to acquire multiple projection images at different angles, which are then mathematically processed to generate multiple reconstructed images representing different slices through the breast tissue. This dimensional change allows lesions to be visualized in their true spatial context, separating them from overlying dense tissue that causes obscuration in 2D images.
2Reliability
If x-ray tomosynthesis imaging is used to improve cancer detection, then detection accuracy improves, but the volume of data to be reviewed increases significantly
Solution Approach 1:
The patent divides the three-dimensional tomosynthesis data into multiple two-dimensional reconstructed image slices (e.g., 40-60 slices) representing different depths through the breast tissue. Each slice can be independently reviewed by the radiologist, allowing systematic examination of the entire volume without being overwhelmed by a single large 3D dataset. This segmentation transforms the review process into a series of manageable 2D image assessments.
3Reliability
If multiple reconstructed images are generated from tomosynthesis data, then cancer detection capability improves, but the time required to review all images increases
Solution Approach 1:
The patent integrates computer-aided detection (CAD) functionality directly into the tomosynthesis review system. The CAD system automatically analyzes all reconstructed slices, generates candidate findings, and presents them to the radiologist with confidence ratings. This multi-functional approach combines automated analysis with human expertise, allowing the radiologist to focus on evaluating CAD-suggested findings rather than manually examining every slice, thereby reducing review time while maintaining high detection accuracy.
4Reliability
If CAD systems are used to assist in reviewing tomosynthesis data, then detection accuracy improves, but CAD markers can be easily overlooked due to the large number of slices
Solution Approach 1:
The patent introduces an interactive user interface that serves as an intermediary between the CAD system and the radiologist. This interface provides graphical navigation tools, including a depth navigation slider that allows the radiologist to quickly jump to specific slices containing CAD markers. The interface also provides visual indicators showing the depth and location of CAD findings within the 3D volume, making it easy to locate and evaluate markers without manually scrolling through all slices. This intermediary layer simplifies the interaction and ensures CAD findings are not overlooked.
Data Source
AI summary
Methods, systems, and related computer program products for processing and displaying computer-aided detection (CAD) information associated with medical breast x-ray images, such as breast x-ray tomosynthesis volumes, are described. An interactive graphical user interface for displaying a tomosynthesis data volume is described that includes a display of a two-dimensional composited image having slabbed sub-images spatially localized to marked CAD findings. Also described is a graphical navigation tool for optimized CAD-assisted viewing of the data volume, comprising a plurality of CAD indicator icons running near and along a slice ruler, each CAD indicator icon spanning a contiguous segment of the slice ruler and corresponding in depthwise position and extent to a subset of image slices spanned by the associated CAD finding, each CAD indicator icon including at least one single-slice highlighting mark indicating a respective image slice containing viewable image information corresponding to the associated CAD finding.