Breast Biopsy Targeting via 3D Microcalcification Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current X-ray guided breast biopsy techniques, such as stereotaxy and digital breast tomosynthesis, are time-consuming and imprecise when targeting microcalcifications, especially when they form multiple clusters, leading to prolonged biopsy durations and potential inaccuracies.
Innovation Solution
A method for processing X-ray images that automatically computes and displays a needle target position within the breast, reducing the time required for target selection by constructing a 3D volume from raw X-ray images, detecting regions of interest like microcalcifications, and calculating optimal needle positions for biopsy, using image processing and graphical interfaces to assist radiologists.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If digital breast tomosynthesis is used to provide 3D representation for biopsy guidance, then measurement precision is improved, but duration of action increases due to time-consuming navigation through 3D volume
Solution Approach 1:
The system performs preliminary detection of microcalcifications and automatic computation of optimal needle target positions before the biopsy procedure. The image processing unit analyzes the 3D volume, identifies regions of interest, and calculates target positions in advance, so that during the actual biopsy, the radiologist only needs to confirm the pre-computed target rather than navigating through the entire 3D volume manually.
Solution Approach 2:
The system enables self-service by automatically detecting microcalcifications and computing optimal biopsy target positions without requiring extensive manual intervention. The image processing unit autonomously analyzes the 3D tomosynthesis data, identifies suspicious regions, and determines needle insertion points, reducing the radiologist's workload from manual navigation to verification.
2Measurement precision
If manual navigation through 3D volume is performed to select target point, then measurement precision is improved, but productivity decreases due to increased time consumption
Solution Approach 1:
The system performs preliminary detection of microcalcifications and automatic computation of optimal needle target positions before the biopsy procedure. The image processing unit analyzes the 3D volume, identifies regions of interest, and calculates target positions in advance, so that during the actual biopsy, the radiologist only needs to confirm the pre-computed target rather than navigating through the entire 3D volume manually.
Solution Approach 2:
The system replaces the mechanical manual navigation process with automated image processing algorithms. Instead of the radiologist manually browsing through 3D slices and mentally localizing targets, the computer automatically processes the 3D data, detects microcalcifications, and computes optimal target positions, thereby increasing productivity while maintaining precision.
3Measurement precision
If radiologist manually selects target point in 3D volume, then measurement precision is improved, but device complexity increases due to need for multiple image acquisition and navigation systems
Solution Approach 1:
The image processing unit performs multiple functions within a single integrated system: it processes the 3D tomosynthesis volume, detects microcalcifications, identifies regions of interest, and computes optimal needle target positions. This multi-functional approach consolidates what would otherwise require separate navigation systems, image analysis tools, and target marking mechanisms into one unified device, reducing overall system complexity.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method of processing X-ray images of a breast (B), the method comprising the steps of - generating (GEN3DV) a 3D volume from a plurality of X-ray images, - processing (DETMC) the 3D volume and/or the plurality of X-ray images, - computing (DETMC-CTR), for each region of interest in the 3D volume, a 3D characteristic position of the region of interest, - computing (CALCTGT) at least one needle target 3D position from a plurality of said computed 3D characteristic positions, - associating (SELIMG-TGT) each needle target position with a target image, - displaying (DISP-IMG) a slice image on a graphical interface, and - if the current slice image is a target image, displaying (DISP-TGT) on the graphical interface a target marker indicating each needle target position associated with the displayed current slice image.