Medical Image Quantification via Length Scale Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for diagnosing pulmonary diseases like emphysema, such as diffusing capacity of the lungs for carbon monoxide and high-resolution computed tomography, fail to provide objective information on the extent and local distribution of emphysema, relying on intuitive interpretation which lacks objectivity and correlation with lung anatomy.
Innovation Solution
A method of quantifying medical images by acquiring regions of interest, filtering their sizes using length scale analysis, and classifying them based on size, allowing for objective visualization and distinction within the image, which includes noise elimination and image processing techniques like Gaussian low-pass filtering and hole filling.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If CT-based LAA% or EI is used to diagnose emphysema, then the left and right lungs are distinctively shown and emphysema distribution is shown at pixel level, but the determination of size and distribution is not objective
Solution Approach 1:
The patent applies length scale analysis to transform the visual assessment of emphysema into quantitative measurements. By analyzing images at multiple length scales and calculating specific parameters (such as cluster size, density, and distribution patterns), the system objectively quantifies emphysema characteristics that were previously only assessable through subjective visual inspection.
Solution Approach 2:
The patent introduces an image processing system as an intermediary between the CT scan and the diagnostic interpretation. This system processes the raw CT images through length scale analysis algorithms, generating objective quantitative data that mediates between the imaging data and the clinical decision-making process, eliminating the need for direct subjective visual assessment.
2Device complexity
If average numerical value (EI) is used to represent emphysema, then the diagnosis is simplified, but sufficient information about correlation between size, local distribution and lung anatomy is lost
Solution Approach 1:
The patent segments the emphysema assessment into multiple quantitative components through length scale analysis. Instead of using a single average value, the system divides the analysis into different spatial scales, identifying and measuring distinct emphysema clusters, their sizes, densities, and distribution patterns. This segmentation preserves detailed information while organizing it into structured quantitative data.
Solution Approach 2:
The patent adds spatial dimensionality to the emphysema assessment by analyzing images at multiple length scales. This transforms the one-dimensional average emphysema index into multi-dimensional quantitative data that captures size, distribution, and spatial relationships, providing comprehensive information without excessive complexity.
3Ease of operation
If DLco is used to measure pulmonary function, then the test is simple and provides average information, but it cannot distinguish left and right lungs or provide location-based information
Solution Approach 1:
The patent applies local quality analysis by examining emphysema characteristics at different locations within the lungs using length scale analysis. The system identifies and quantifies emphysema clusters in specific lung regions, providing location-based information that complements the global average measurements from DLco testing.
Solution Approach 2:
The patent adds spatial localization as an additional dimension to pulmonary function assessment. By combining the functional data from DLco with the spatially-resolved structural data from length scale analysis of CT images, the system provides both average and location-specific information in an integrated framework.
Data Source
AI summary
A method of quantifying a medical image is disclosed herein. The method of quantifying a medical image includes acquiring regions of interest based on a medical image; filtering the sizes of the acquired regions of interest using length scale analysis; classifying the regions of interest whose sizes have been filtered, according to the sizes of the regions of interest; and visualizing the regions of interest whose sizes have been filtered, so that the regions of interest whose sizes have been filtered are distinguished from each other in the medical image according to the sizes of the regions of interest.