COPD CT Visualization Using Inspiratory-Expiratory Voxel Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for diagnosing chronic obstructive pulmonary disease (COPD) from medical images lack sufficient correlation with clinical pulmonary function testing (PFT) results, leading to inaccurate classification of lung regions and a need for improved quantitative assessment and visualization.
Innovation Solution
A visualization method and apparatus that classifies lung regions into functional air trapping (fAT), emphysema, and normal regions by analyzing intensity differences in inspiratory and expiratory CT images using threshold values, providing quantitative assessment information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If prior art methods classify lung regions based on CT intensity values, then classification can be performed, but the correlation with clinical PFT results is not high enough
Solution Approach 1:
The patent applies parameter changes by transitioning from single-phase CT intensity value classification to dual-phase (inspiratory and expiratory) intensity value classification. This involves calculating both absolute intensity values and differential changes between phases, thereby improving the correlation with PFT results and diagnostic accuracy for COPD assessment
2Reliability
If normal regions are classified as disease regions, then sensitivity increases, but specificity decreases leading to false positives
Solution Approach 1:
The patent applies local quality by establishing different classification criteria and threshold values for different lung regions based on their local characteristics. Normal regions are distinguished from disease regions through region-specific intensity value ranges and differential change thresholds, preventing false positives while maintaining detection sensitivity
3Device complexity
If emphysema and fAT regions are distinguished using single threshold values, then classification is simple, but functional overlapping regions are misclassified
Solution Approach 1:
The patent applies segmentation by dividing the classification process into multiple stages: first classifying based on inspiratory intensity values, then applying expiratory differential changes as a second classification criterion. This multi-stage segmentation approach accurately identifies functional overlapping regions without excessively increasing system complexity
Data Source
AI summary
Disclosed are a visualization method for assisting medical image diagnosis comprising: acquiring first intensity values of first voxels in a lung region during inspiration, segmented from a chest computed tomography (CT) image acquired during inspiration, as first coordinate values of the first voxels; acquiring differences between second intensity values of second voxels, registered into the first voxels as voxels in the lung region during expiration segmented from a chest CT image acquired during expiration, and the first intensity values as second coordinate values of the first voxels; and visualizing a distribution of the first voxels by mapping the first voxels based on the first coordinate values and the second coordinate values.