Lesion Positioning via Nipple Distance Tomogram Composite
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for identifying lesions in medical images, particularly from MRI to ultrasonic tomograms, are cumbersome due to the need for complex deformation simulations and extensive measurement input, and fail to accurately locate lesions on arbitrary slices, increasing operator load and risk of incorrect association.
Innovation Solution
An information processing system that calculates and displays support information by using the consistent three-dimensional distance from the nipple to a lesion, regardless of body posture, allowing for simplified image analysis and reduced operator input, by generating composite images with arcs indicating the likelihood of lesion presence on ultrasonic tomograms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If deformation simulation calculation is performed to estimate lesion position, then position estimation accuracy is improved, but calculation time and operator waiting time increase
Solution Approach 1:
The patent extracts only the essential geometric relationship (distance from nipple to lesion) needed for position estimation, discarding the complex deformation simulation process. By taking out just the critical measurement (nipple-lesion distance) and applying it directly to the target image, the system achieves position estimation without time-consuming calculations.
Solution Approach 2:
The patent creates a simple geometric copy of the lesion position relationship from the source image (distance from nipple) and applies it to the target image. Instead of simulating complex tissue deformation, it copies the essential spatial relationship and reproduces it in the new context, achieving fast results with sufficient accuracy.
2Measurement precision
If multiple measurement information inputs are required for deformation simulation, then estimation accuracy is improved, but operator workload and complexity increase
Solution Approach 1:
The patent extracts only the single essential measurement (distance from nipple to lesion) that is sufficient for position estimation, eliminating the need for multiple complex measurements. This reduces operator input burden while maintaining practical accuracy for the clinical task.
Solution Approach 2:
The system uses the nipple as a self-identifying reference point that is naturally present in breast images, eliminating the need for operators to manually set up complex coordinate systems or reference frameworks. The nipple automatically serves as the origin for distance measurement.
3Reliability
If complex deformation simulation is used to account for posture changes, then position accuracy across different postures is improved, but device complexity and computational requirements increase
Solution Approach 1:
The patent changes the approach from simulating complex tissue deformation parameters to using a simple distance parameter that remains relatively stable across postures. By focusing on the invariant nipple-lesion distance rather than variable tissue deformation parameters, the system achieves cross-posture accuracy with minimal computational complexity.
4Measurement precision
If only slice-specific region information is provided, then localization precision for that slice is improved, but ability to locate lesions on arbitrary slices is reduced
Solution Approach 1:
The patent transitions from providing only two-dimensional slice-specific region information to providing three-dimensional spatial information (distance from nipple in 3D space). This allows the same distance measurement to be applied to any slice containing the lesion, as the 3D distance relationship holds across all slices regardless of orientation or position.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
A tomogram of an object is acquired. A place in a tomogram which corresponds to a portion spaced apart from a reference point in the object by a predetermined distance is specified. A composite image is generated by combining the tomogram with information indicating the specified place. The composite image is output.