Medical Image Region Extraction via Phase Estimation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing medical image processing techniques face challenges in accurately extracting anatomical regions from images with uncertain or erroneous phase information, leading to decreased accuracy in region extraction and parameter selection.

Innovation Solution

An image processing apparatus that estimates the phase of medical images by calculating gray level distribution information from selected anatomical portions with minimal positional variations, allowing for accurate prior distribution estimation and improved region extraction even without phase information.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If phase information is used for region extraction, then extraction accuracy is improved, but the system becomes vulnerable to errors when phase information is deficient or incorrect

Engineering Contradiction:
Improveregion extraction accuracyVSAvoidsystem robustness to phase information errors
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent introduces an intermediary phase estimation process that mediates between the input image and the region extraction process. Instead of directly using potentially erroneous phase information, the system estimates the phase from image data and uses this estimation as an intermediary step to select appropriate processing parameters, thereby improving reliability while maintaining accuracy

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements a feedback mechanism where the estimated phase information is used to adjust processing parameters, and the results are evaluated to refine the estimation. This closed-loop approach allows the system to compensate for errors in initial phase information and improve both accuracy and robustness

Inventive Principle:
Principle #23Feedback

2Ease of operation

If one image processing parameter is used for all phases, then processing simplicity is maintained, but extraction accuracy decreases for specific anatomical portions

Engineering Contradiction:
Improveprocessing simplicityVSAvoidanatomical portion extraction accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent applies dynamics by making the image processing parameters adaptive rather than static. The system dynamically selects processing parameters based on the estimated phase of the input image, allowing the parameters to change according to the specific anatomical portion and phase characteristics, thereby maintaining both simplicity and accuracy

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes processing parameters based on the estimated phase information. Different parameters are applied for different phases (e.g., arterial phase vs. venous phase) to optimize extraction accuracy for specific anatomical portions, while the automatic parameter selection maintains operational simplicity

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If prior distribution of gray levels is accurately provided, then region extraction accuracy is improved, but the system requires reliable phase information which may not be available

Engineering Contradiction:
Improveregion extraction accuracyVSAvoidphase information availability
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent performs preliminary phase estimation from the input image data before proceeding to region extraction. This preliminary action of estimating the phase allows the system to prepare the appropriate prior distribution of gray levels in advance, ensuring accurate extraction even when original phase information is missing or erroneous

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system uses the input image data itself to estimate the phase information needed for accurate processing. By self-generating the phase estimation from the available image data, the system does not rely on external phase information that may be deficient or erroneous, thereby maintaining extraction accuracy without requiring additional reliable information

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP3150125B1Image processing apparatus, method of controlling image processing apparatus, and storage medium
Publication Date: 2021.03.10 CANON KK
  • EP3150125B1 patent drawingFigure 1
  • EP3150125B1 patent drawingFigure 2
  • EP3150125B1 patent drawingFigure 3

AI summary

An image processing apparatus comprises obtainment means (1010) for obtaining a contrast material-enhanced image of an object; first region extraction means (1040) for extracting a first region representing a first anatomical portion of the object from the image; estimation means (1060) for estimating a state of the image concerning a temporal change in gray level from the first region; and second region extraction means (1070) for extracting a second region representing a second anatomical portion of the object from the image based on an estimation result obtained by the estimation means.