HRCT Lung Tissue Texture Analysis for UIP Differentiation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

High-resolution computed tomography (HRCT) imaging struggles to reliably distinguish between interstitial lung diseases (ILDs) due to similar manifestations, leading to false alarms and missed diagnoses, necessitating an automated system to differentiate between usual interstitial pneumonia (UIP) and non-UIP tissue regions.

Innovation Solution

A method involving the extraction of textural and localization features from HRCT scans, using first-order statistics, second-order statistics, gray level co-occurrence matrices, gray level run-length matrices, and location features, followed by feature selection and construction of a predictive model to accurately identify UIP voxels, enabling voxel-wise prediction of UIP presence.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If high-resolution computed tomography (HRCT) imaging is used to detect interstitial lung diseases, then the imaging resolution and detail are improved, but the ability to reliably distinguish between different ILDs deteriorates due to similar manifestations

Engineering Contradiction:

Improveimaging resolutionVSAvoiddisease differentiation accuracy

Core Design Contradiction:

Measurement precisionVSReliability

Solution Approach 1:

The patent segments the lung tissue into multiple regions of interest (ROIs) based on anatomical location and disease patterns. By dividing the lung into distinct zones and analyzing texture features independently in each region, the system can identify location-specific patterns that help differentiate between ILD types that appear similar in overall HRCT images.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent transitions from analyzing only image intensity values to extracting multi-dimensional texture features including spatial relationships, gray-level co-occurrence patterns, and run-length characteristics. This dimensional expansion allows the system to capture subtle structural differences between diseases that are not apparent in standard HRCT visualization.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

2Reliability

If manual review of HRCT images is performed to distinguish between ILDs, then diagnostic accuracy may be improved, but the time consumption and workload increase significantly

Engineering Contradiction:

Improvediagnostic accuracyVSAvoiddiagnosis time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent implements an automated diagnostic support system that performs feature extraction, texture analysis, and disease pattern recognition without requiring manual intervention. The system processes HRCT images automatically, generating diagnostic recommendations that assist radiologists in quickly differentiating between ILD types while maintaining high accuracy.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces manual visual inspection and subjective interpretation with computer-based automated analysis. Machine learning algorithms and texture feature extraction methods substitute for the mechanical process of manual image review, enabling rapid objective assessment of lung tissue patterns.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Reliability

If invasive procedures are performed to confirm ILD diagnoses, then diagnostic certainty is improved, but patient quality of life deteriorates due to unnecessary procedures

Engineering Contradiction:

Improvediagnosis certaintyVSAvoidpatient quality of life

Core Design Contradiction:

ReliabilityVSObject-affected harmful factors

Solution Approach 1:

The patent performs comprehensive automated texture analysis and pattern recognition on HRCT images before patients undergo invasive procedures. By providing a preliminary automated diagnosis with high confidence levels, the system can identify cases where invasive procedures are unnecessary, thereby preventing harm to patients while maintaining diagnostic certainty for those who truly need further intervention.

Inventive Principle:

Principle #10Preliminary action

4Adaptability or versatility

If multiple ILD patterns are detected automatically, then the coverage of detectable diseases is improved, but the complexity of the diagnostic system increases

Engineering Contradiction:

Improvedisease detection coverageVSAvoidsystem complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent develops a universal automated analysis platform that can detect and differentiate multiple types of interstitial lung diseases using a single integrated system. The same core architecture processes various ILD patterns by analyzing different texture feature combinations, eliminating the need for separate specialized systems for each disease type.

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS9852501B2Textural analysis of diffused disease in the lung

Publication Date: 2017.12.26 GE PRECISION HEALTHCARE LLC

AI summary

A method of creating a diagnostic evaluation for usual interstitial pneumonia is provided, including obtaining a first plurality of series of HRCT lung slices indicating the presence of UIP, obtaining an identification of UIP and non-UIP voxels, extracting textural and localization features from the UIP and non-UIP voxels, selecting features that are more accurate in differentiating UIP voxels from non-UIP voxels than other features are, eliminating features highly correlated with a more accurate feature, and constructing a predictive model by performing a second classifier to provide a probability that a voxel signifies the presence of UIP. Also provided is a method of identifying UIP in a subject's lung by applying a diagnostic evaluation for UIP that was created with the foregoing method.