IVUS Diagnosis Support With Tissue Classification for Accurate 3D Imaging

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current IVUS methods face challenges in accurately generating three-dimensional images of biological tissues due to high blood cell noise interference and the difficulty in differentiating tissue regions from blood cells, leading to inaccurate structural representations that can compromise surgical safety, especially for inexperienced operators.

Innovation Solution

A diagnosis support device that classifies pixels in a two-dimensional IVUS image into multiple classes, including biological tissue and blood cell classes, and generates a three-dimensional image by excluding blood cell pixels, allowing for real-time differentiation of biological tissue and medical devices, and displays these in distinguishable formats.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If a three-dimensional image is automatically generated from a two-dimensional IVUS image, then the ease of operation is improved for inexperienced doctors, but the measurement precision deteriorates due to blood cell noise interference

Engineering Contradiction:
Improveease of operationVSAvoidmeasurement precision
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent extracts and removes blood cell pixels from the IVUS image data before three-dimensional image generation. The control unit identifies pixels corresponding to blood cells based on their echo characteristics and excludes them from the tissue structure reconstruction, thereby eliminating the harmful noise while preserving the useful tissue information for accurate 3D rendering.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent introduces an intermediary classification process that acts as a mediator between the raw IVUS image and the final three-dimensional image. This intermediate step involves classifying each pixel into different types (blood cell, tissue, medical device) before reconstruction, serving as a filter that cleans the data while maintaining the operational simplicity for users.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If pixel classification into multiple classes is performed, then the measurement precision of biological tissue structure is improved, but the device complexity increases

Engineering Contradiction:
Improvemeasurement precisionVSAvoiddevice complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the pixel classification task into distinct categories: blood cell pixels, biological tissue pixels, and medical device pixels. By dividing the complex image processing into these manageable segments with specific classification criteria for each type, the system achieves high measurement precision while keeping the complexity organized and controllable through structured processing stages.

Inventive Principle:
Principle #1Segmentation

3Manufacturing precision

If blood cell pixels are excluded from three-dimensional image generation, then the manufacturing precision of tissue structure representation is improved, but the loss of information increases

Engineering Contradiction:
Improvemanufacturing precisionVSAvoidloss of information
Core Design Contradiction:
Manufacturing precisionVSLoss of information

Solution Approach 1:

The patent selectively extracts only the harmful blood cell pixels while preserving all useful information from tissue and medical device pixels. The classification system ensures that only pixels corresponding to blood cells are removed, while maintaining complete information about the biological tissue structure and medical devices for accurate three-dimensional reconstruction.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent converts the harmful blood cell noise into a beneficial classification target. By identifying blood cell pixels as a distinct class with specific echo characteristics, the system uses what was originally harmful information as a criterion for exclusion, thereby improving the final image quality while maintaining awareness of the removed data.

Inventive Principle:
Principle #22Blessing in disguise (Convert harm into benefit)

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

Improves the accuracy of three-dimensional tissue structure representation, enabling safer and more precise medical procedures by reducing operator fatigue and minimizing procedural errors through real-time, accurate three-dimensional imaging.

Implementation Method 1

a probe that transmits the ultrasound inside the biological tissue and inputs the signal of the reflected wave

Methodology Applied
Scientific EffectUltrasound: Ultrasound

Implementation Method 2

a probe that transmits the ultrasound inside the biological tissue and inputs the signal of the reflected wave

Methodology Applied
Scientific EffectReflection: Reflection

Data Source

PatentEP3949866B1Diagnosis support device, diagnosis support system, and diagnosis support method
Publication Date: 2025.07.09 TERUMO KK
  • EP3949866B1 patent drawingFigure 1
  • EP3949866B1 patent drawingFigure 2
  • EP3949866B1 patent drawingFigure 3

AI summary

A diagnosis support device includes a control unit that correlates a plurality of pixels included in a two-dimensional image with two or more classes including a biological tissue class, the two-dimensional image being generated by using a signal of a reflected wave of ultrasound transmitted inside a biological tissue through which blood passes and the two-dimensional image including the biological tissue, generates a three-dimensional image of the biological tissue from a pixel group correlated with the biological tissue class, and performs control of displaying the generated three-dimensional image of the biological tissue.