3D Airway and Pulmonary Lobe Segmentation Using Directionality Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for segmenting airways and pulmonary lobes from 3D images of the human body face challenges due to partial volume effects, noise, and decreased contrast in CT images, particularly for bronchioles with diameters less than 2 mm and thin fissures between pulmonary lobes, which hinders accurate disease analysis and surgical planning.
Innovation Solution
A method and apparatus that utilize region growing techniques, noise removal, and anatomical directionality of signal intensity changes to segment airways and pulmonary lobes, allowing for automated segmentation without manual input, and enable precise identification of bronchioles and fissures using 3D image processing and analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional airway segmentation methods are used, then the process is simple, but segmentation accuracy deteriorates for bronchioles with diameter less than 2 mm due to partial volume effect and noise
Solution Approach 1:
The patent divides airway segmentation into multiple stages: initial segmentation to obtain candidate regions, refinement segmentation to remove noise, and validation to ensure accuracy. This multi-stage approach improves segmentation precision for small bronchioles while managing complexity through systematic breakdown of the segmentation process
Solution Approach 2:
The patent employs 3D image processing and analyzes airways in multiple spatial dimensions simultaneously. By considering volumetric data and spatial relationships in three dimensions, the method improves detection accuracy of small-diameter bronchioles that are difficult to resolve in conventional 2D or single-dimension approaches
2Extent of automation
If manual designation of bronchus is required, then segmentation can be guided, but automation level deteriorates requiring user input
Solution Approach 1:
The system performs automated airway segmentation by itself without requiring manual bronchus designation. The algorithm automatically identifies airway structures, segments candidate regions, and refines results through noise removal, achieving both high automation and maintained segmentation accuracy
Solution Approach 2:
The patent implements iterative refinement where segmentation results are evaluated and fed back into the process. Candidate regions are segmented, noise is removed based on similarity criteria, and results are validated, creating a feedback loop that maintains accuracy while achieving automation
3Measurement precision
If thin-film fissure less than 1 mm is used to separate pulmonary lobes, then anatomical separation is accurate, but detection difficulty increases due to low contrast and noise
Solution Approach 1:
The patent performs preliminary segmentation of the lung region before attempting to detect fissures. By first establishing the overall lung boundaries and internal structure, the system prepares the data in advance, making subsequent thin-fissure detection easier and more accurate
Solution Approach 2:
The patent uses 3D image processing to detect fissures, analyzing the thin-film structures in three-dimensional space. This volumetric approach enhances the detectability of sub-1mm fissures by considering their spatial context and relationships across multiple dimensions, overcoming the limitations of 2D detection
4Object-affected harmful factors
If low-dose CT is used, then radiation dose is reduced, but image quality deteriorates with increased noise and decreased contrast
Solution Approach 1:
The patent employs noise removal algorithms that convert the harmful noise in low-dose CT images into beneficial signal enhancement. By identifying and removing noise based on similarity criteria while preserving true anatomical structures, the system recovers image quality despite the reduced radiation dose
Solution Approach 2:
The patent applies image processing parameters and algorithms to enhance low-dose CT images. By adjusting segmentation thresholds, noise filters, and contrast enhancement parameters, the system optimizes image quality for diagnostic purposes while maintaining the benefits of low radiation exposure
Data Source
AI summary
Provided is a method and apparatus for segmenting airways and pulmonary lobes. An image processing apparatus obtains a first candidate region of an airway from a three-dimensional (3D) human body image by using a region growing method, obtains a second candidate region of the airway based on a directionality of a change in signal intensity of voxels belonging to a lung region segmented from the 3D human body image, segments an airway region by removing noise based on similarity of a directionality of a change in signal intensity of voxels belonging to a third candidate region acquired by combining together the first and second candidate regions. Furthermore, the image processing apparatus segments a lung region from a 3D human body image by using a region growing method, obtains a fissure candidate group between pulmonary lobes based on a directionality of a change in signal intensity of voxels belonging to the lung region, reconstructs an image of the lung region including the fissure candidate group into an image viewed from a front side of a human body and generates a virtual fissure based on a fissure candidate group shown in the reconstructed image, and segments the pulmonary lobes by using the virtual fissure.