Lung Nodule Risk Characterization via Perinodular Radiomics

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional methods for diagnosing lung nodules using CT imagery struggle to accurately distinguish between benign and malignant nodules, leading to unnecessary invasive procedures and high costs, due to subjective and time-consuming visual characterization by human pathologists, which results in variability and inefficiency.

Innovation Solution

The use of computerized methods that analyze tortuosity features extracted from the perinodular region associated with a nodule, employing techniques such as spectral embedding gradient vector flow active contour segmentation and principal component analysis-variable importance projection, to differentiate between benign and malignant nodules, reducing reliance on invasive procedures.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional visual characterization by human pathologists is used, then diagnostic assessment can be performed, but the process is time-consuming and subjective with inter-rater and intra-rater variability

Engineering Contradiction:

Improvediagnostic accuracyVSAvoidtime consumption

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent replaces the mechanical visual inspection process performed by human pathologists with an automated image processing system that uses algorithms to extract and analyze morphological features of lung nodules, thereby eliminating human subjectivity and time consumption while maintaining or improving diagnostic accuracy

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-service diagnosis by automatically processing CT images, extracting features, and generating diagnostic assessments without requiring human pathologist intervention for the actual measurement and analysis, though human oversight may still be involved in final interpretation

Inventive Principle:

Principle #25Self-service

2Reliability

If conventional CT-based approaches are used to distinguish benign granuloma from malignant adenocarcinoma, then imaging can be performed, but reliable discrimination is difficult or impossible

Engineering Contradiction:

Improvediagnostic reliabilityVSAvoidnodule characterization accuracy

Core Design Contradiction:

ReliabilityVSMeasurement precision

Solution Approach 1:

The patent segments the lung nodule into distinct morphological components and analyzes specific features such as spiculation, lobulation, and texture patterns that are not visible through conventional visual inspection, enabling more precise differentiation between benign and malignant nodules

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system transforms the diagnostic approach by changing from subjective visual assessment to quantitative measurement of multiple morphological parameters, including but not limited to spiculation index, lobulation index, and texture features, which provides more reliable and precise discrimination capability

Inventive Principle:

Principle #35Parameter changes

3Measurement precision

If invasive procedures such as surgical resections and biopsies are performed to diagnose nodules, then definitive diagnosis can be obtained, but patients face additional risks and costs increase

Engineering Contradiction:

Improvediagnostic accuracyVSAvoidpatient risk

Core Design Contradiction:

Measurement precisionVSObject-affected harmful factors

Solution Approach 1:

The patent introduces an intermediate diagnostic step using automated morphological analysis of CT images that can provide sufficient diagnostic information to avoid unnecessary invasive procedures, serving as a mediator between initial nodule detection and invasive diagnostic intervention

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system creates a detailed quantitative copy of the nodule's morphological characteristics from the CT image, allowing virtual examination and diagnosis without physical intervention, thereby eliminating the need for invasive procedures in cases where the automated analysis provides confident classification

Inventive Principle:

Principle #26Copying

Data Source

PatentUS10470734B2Characterizing lung nodule risk with quantitative nodule and perinodular radiomics

Publication Date: 2019.11.12 CASE WESTERN RESERVE UNIV

AI summary

Embodiments associated with classifying a region of tissue using features extracted from nodules and surrounding structures. One example apparatus includes a feature extraction circuit configured to automatically extract a first set of quantitative features from a nodule represented in at least one CT image, and automatically extract a second set of quantitative features from the lung parenchyma region immediately surrounding the nodule represented in the at least one CT image; a feature selection circuit configured to select an optimally predictive feature set from the first set of quantitative features and the second set of quantitative features; and a training circuit configured to train a classifier using the optimally predictive feature set to assign malignancy risk to a lung nodule represented in a CT image of a region of tissue demonstrating lung nodules. A prognosis or treatment plan may be provided based on the malignancy risk.