Breast Tomosynthesis Lesion Localization via Shape Mapping

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Diagnosing practitioners face challenges in accurately localizing lesions in tomosynthesis image data records of compressed breasts due to the need to estimate non-compressed breast shapes from compressed images, leading to increased mental effort and risk of errors.

Innovation Solution

A method that automatically generates spatial descriptors and pictograms by mapping compressed breast shapes to non-compressed shapes using a combination of data-driven and biomechanical simulations, with thin plate spline transformations, to provide accurate and consistent position information for target structures.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If spatial descriptors are specified based on compressed breast images, then documentation can be completed using available imaging data, but the accuracy of lesion localization deteriorates because the descriptors do not accurately represent the non-compressed breast anatomy

Engineering Contradiction:

Improvedocumentation efficiencyVSAvoidlesion localization accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The system performs preliminary estimation of the non-compressed breast shape and pre-calculates the mapping between compressed and non-compressed coordinate systems before final lesion localization. This allows spatial descriptors to be accurately transformed from compressed image coordinates to non-compressed breast anatomy coordinates, resolving the contradiction by preparing the transformation framework in advance.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent introduces an intermediary computational model that estimates the non-compressed breast shape and serves as a bridge between compressed image coordinates and non-compressed anatomical coordinates. This intermediary transformation model allows accurate mapping of lesion positions without requiring direct visualization of the non-compressed breast, thus maintaining both efficiency and accuracy.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Measurement precision

If practitioners manually estimate non-compressed breast shapes from compressed images, then accurate spatial descriptors can be obtained, but the mental effort and time required increase significantly

Engineering Contradiction:

Improvespatial descriptor accuracyVSAvoidtime for lesion documentation

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system performs automatic estimation of the non-compressed breast shape and automatic calculation of coordinate transformations without requiring practitioner intervention. The computational model self-adjusts based on the compression parameters and image data, eliminating the need for manual estimation while maintaining high accuracy in spatial descriptor generation.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces the manual cognitive process of estimating non-compressed breast shapes with an automated computational model. The system uses algorithms to estimate breast shape and calculate coordinate transformations, substituting the practitioner's mental estimation task with automated image processing and mathematical transformation, thus reducing time loss while maintaining accuracy.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Measurement precision

If comprehensive coordinate transformations are performed to map lesion positions from compressed to non-compressed breast anatomy, then localization accuracy improves, but the computational complexity and processing time increase

Engineering Contradiction:

Improvelesion position accuracyVSAvoidtransformation system complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The coordinate transformation process is segmented into distinct computational steps: first estimating the non-compressed breast shape from compression parameters, then calculating the mapping between compressed and non-compressed coordinate systems, and finally applying the transformation to lesion positions. This segmentation allows each step to be optimized independently and facilitates efficient implementation of the overall transformation pipeline.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS9962129B2Method and apparatuses for assisting a diagnosing practitioner with describing the location of a target structure in a breast

Publication Date: 2018.05.08 SIEMENS HEALTHINEERS AG

AI summary

An operating method assists a diagnosing practitioner with describing the location of a target structure in a tomosynthesis image data record of a compressed breast of a patient. The target structure is localized by a first spatial information item. The method includes: a) ascertaining a first shape information item describing at least one first, compressed breast shape in the tomosynthesis image data record, b) determining a second shape information item describing the breast in a non-compressed breast shape, from the first shape information item, c) mapping the position of the target structure from the compressed breast shape and to the non-compressed breast shape using at least the second shape information item for ascertaining a second spatial information item relating to the non-compressed breast shape, and d) transforming the second spatial information item into a pictogram information item facilitating an abstracted pictorial representation and/or describing the latter.