Joint organ segmentation and characteristics estimation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current image segmentation methods in medical imaging, such as those for cardiac and retinal images, are tedious, time-consuming, and prone to inter and intra-observer variability, especially when performed manually, and lack an efficient automated approach that effectively leverages the interdependency between organ segmentation and characteristics estimation.

Innovation Solution

A computer-implemented method and system that jointly determines image segmentation and characterization by using a feedback loop to interweave organ segmentation and characteristics estimation, training models to predict organ characteristics and segmentations iteratively, and employing an ensemble of shape regressors to improve both processes simultaneously.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual segmentation is performed, then segmentation can be done with human judgment, but it is tedious and time-consuming

Engineering Contradiction:

Improvesegmentation accuracyVSAvoidsegmentation time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent replaces the manual mechanical segmentation process with an automated computer-based system that uses machine learning models and image processing algorithms to perform segmentation and characteristics estimation, eliminating the need for manual human operation while maintaining or improving accuracy

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-service automation where the computer automatically performs segmentation and characteristics estimation without requiring human intervention during the actual segmentation process, making the system independent and efficient

Inventive Principle:

Principle #25Self-service

2Ease of operation

If manual segmentation is performed, then segmentation can be done, but it suffers from inter and intra-observer variability

Engineering Contradiction:

Improvesegmentation flexibilityVSAvoidsegmentation consistency

Core Design Contradiction:

Ease of operationVSReliability

Solution Approach 1:

The patent replaces the human-operated segmentation process with an automated computer-based system that applies consistent algorithms and criteria, eliminating the variability inherent in manual human judgment while maintaining operational flexibility through programmable parameters

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Ease of manufacture

If separate processing is used for segmentation and characteristics estimation, then each task can be optimized independently, but the interdependency between them is not leveraged

Engineering Contradiction:

Improvemodel training simplicityVSAvoidjoint estimation accuracy

Core Design Contradiction:

Ease of manufactureVSManufacturing precision

Solution Approach 1:

The patent merges the segmentation and characteristics estimation tasks into a unified joint processing framework where both tasks are performed simultaneously using shared computational resources and interconnected models, leveraging their interdependency to improve overall accuracy

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The system employs multi-functional computational models that can perform both segmentation and characteristics estimation functions, allowing the same system to handle multiple related tasks efficiently while exploiting the relationships between them

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS10229493B2Joint segmentation and characteristics estimation in medical images

Publication Date: 2019.03.12 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

Jointly determining image segmentation and characterization. A computer-generated image of an organ may be received. Organ characteristics estimation may be performed to predict the organ characteristics considering organ segmentation. Organ segmentation may be performed to delineate the organ in the image considering the organ characteristics. A feedback loop feeds the organ characteristics estimation to determine the organ segmentation, and feeds back the organ segmentation to determine the organ characteristics estimation.