3D Ultrasound Acetabulum Segmentation via Graph Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for diagnosing developmental dysplasia of the hip (DDH) using 3D ultrasound face challenges due to complex shape deformities and high inter-observer and inter-scan variability, particularly with 2D ultrasound techniques, and lack of effective automatic segmentation techniques for 3D ultrasound data, which are hindered by noise and artifacts like speckle noise.
Innovation Solution
A semi-automatic segmentation system using a graph search algorithm to identify seed points and interpolate boundary contours over a 3D volume, calculating the acetabular contact angle (ACA) based on a polygonal mesh representation, which reduces variability and noise resilience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If 2D ultrasound images are used for diagnosing DDH, then the diagnostic process is simple, but inter-observer and inter-scan variability is high
Solution Approach 1:
The patent transitions from 2D ultrasound images to 3D surface models of the acetabulum. By constructing three-dimensional surface representations from multiple 2D ultrasound slices, the system enables measurements that are independent of scanning angle and plane orientation, thereby eliminating inter-observer and inter-scan variability while maintaining operational simplicity.
2Productivity
If automatic segmentation techniques are applied to 3D ultrasound data, then processing efficiency is improved, but noise and artifacts like speckle noise make segmentation difficult
Solution Approach 1:
The patent replaces traditional threshold-based segmentation methods with a surface modeling approach that fits analytical surface equations to the ultrasound data. This substitution allows the system to automatically segment the acetabulum in 3D ultrasound volumes while being robust to speckle noise and acoustic shadowing, as the surface fitting process inherently filters out high-frequency noise.
3Manufacturing precision
If manual segmentation is used for complex anatomical structures like the acetabulum, then segmentation accuracy can be maintained, but the process is tedious and time consuming
Solution Approach 1:
The patent implements an automated surface modeling system that performs acetabular segmentation without requiring manual intervention. The system automatically extracts the acetabular surface from 3D ultrasound data by fitting mathematical surface models, eliminating the need for tedious manual tracing while maintaining high segmentation accuracy comparable to or better than expert manual segmentation.
4Extent of automation
If CT imaging is used for hip segmentation, then automatic segmentation can be achieved, but radiation exposure is introduced
Solution Approach 1:
The patent creates accurate 3D surface models of the acetabulum using ultrasound imaging, which serves as a safe alternative to CT scanning. By processing ultrasound data through surface modeling algorithms, the system achieves automatic segmentation capability similar to CT-based methods but without exposing patients to ionizing radiation, making it suitable for pediatric patients who require repeated imaging.
Data Source
AI summary
An anatomical structure is segmented from images generated in an anatomical scan. A user input device is configured for manually identifying two or more seed points on a plurality of the images. An image processor is configured to identify an optimal path through seed points on the plurality of images with a graph search. The optimal path between corresponding seed points on different images defines a boundary contour, and the anatomical structure includes two or more boundary contour. The image processor is further configured to interpolate the boundary contours over a three-dimensional volume using cardinal splines.