Medical Image Segmentation Parallel Processing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Clinicians face difficulties in efficiently controlling segmentation methods during routine clinical practice due to the time-consuming nature of existing user interfaces for medical image processing, which require significant involvement in selecting and fine-tuning segmentation algorithms and parameters.

Innovation Solution

A system and method that allows clinicians to simultaneously apply multiple segmentation methods to a region of interest, displaying the resulting segmentation outputs for easy comparison and selection, thereby reducing the need for extensive technical knowledge and iterative processes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a clinician manually controls segmentation methods by adjusting parameters and selecting algorithms, then the segmentation can be fine-tuned to particular clinical applications, but the process becomes time-consuming and difficult to control during routine clinical practice

Engineering Contradiction:
Improvesegmentation accuracyVSAvoidtime for segmentation control
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system performs multiple segmentation methods in parallel before the clinician needs to make a selection. By pre-computing multiple segmentation results with different algorithms and parameters, the system eliminates the need for the clinician to iteratively adjust parameters and wait for results, thus reducing time loss while maintaining the ability to select the most accurate segmentation for the specific clinical application

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system divides the segmentation task into multiple independent segmentation methods that can be executed separately and simultaneously. Each segmentation method processes the medical image independently, producing distinct segmentation results that are then presented to the clinician for selection, thereby enabling parallel processing and reducing overall processing time

Inventive Principle:
Principle #1Segmentation

2Adaptability or versatility

If multiple segmentation methods are applied to compare results, then the clinician can select the most suitable method, but the complexity of the system increases

Engineering Contradiction:
Improveselection of segmentation methodVSAvoidsystem complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system introduces an intermediary component that manages multiple segmentation methods and presents their results to the clinician in a unified, standardized interface. This intermediary layer handles the complexity of coordinating multiple algorithms and parameters, while presenting a simplified view to the clinician, thus enabling method selection without exposing the full system complexity

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system designs a universal segmentation framework that can accommodate multiple different segmentation algorithms and parameter sets through a common interface and processing architecture. This multi-functional design allows the system to apply various segmentation methods without requiring separate systems for each method, managing complexity through consolidation rather than proliferation

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP2724317B1System and method for processing a medical image
Publication Date: 2018.08.29 KONINKLIJKE PHILIPS NV
  • EP2724317B1 patent drawingFigure 1~2
  • EP2724317B1 patent drawingFigure 3a~3b
  • EP2724317B1 patent drawingFigure 4~5

AI summary

A system (100) for processing a medical image (102), the system being arranged for establishing a region of interest 104 in the medical image, and the system comprising segmentation means (120) for applying a plurality of different segmentation methods (124) to the region of interest for obtaining an associated plurality of segmentation results (122), visualization means (140) for simultaneously displaying the plurality of segmentation results to a user, and a user input (160) for receiving from the user a selection command (162) indicative of a selection of one of the plurality of segmentation results for establishing an associated one of the plurality of different segmentation methods as a selected segmentation method (164).