2D X-Ray Segmentation During Cone Beam CT Spin for Faster 3D Results
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for automatic segmentation of anatomical structures in 3-D image volumes are slower and less accurate than in 2-D image planes, and 2-D segmentation is less reliable due to reduced information content.
Innovation Solution
A method for automatically segmenting anatomical structures using 2-D x-ray images during cone beam CT imaging, allowing processing to begin before the spin is completed, thereby combining the speed of 2-D segmentation with the accuracy of 3-D segmentation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If 3-D image volume segmentation is used, then segmentation accuracy is improved, but processing time increases and speed deteriorates
Solution Approach 1:
The patent divides the 3-D volume segmentation task into multiple 2-D slice segmentation tasks. Each slice is processed independently using 2-D segmentation algorithms, which are computationally faster. The slices are then reconstructed to form the complete 3-D segmentation result, thus achieving fast processing while maintaining accuracy through the cumulative information from multiple slices.
Solution Approach 2:
The patent transforms the 3-D volume segmentation problem into a series of 2-D slice segmentation problems. By processing images in 2-D planes rather than the full 3-D volume, the computational complexity is reduced significantly, enabling faster processing speeds while the stacked 2-D slices collectively provide sufficient information for accurate 3-D structure reconstruction.
2Speed
If 2-D image plane segmentation is used, then processing speed is improved, but segmentation reliability deteriorates due to less information
Solution Approach 1:
The patent processes multiple 2-D slices through the volume rather than a single 2-D image. By segmenting numerous sequential slices and reconstructing the 3-D structure from these segments, the system accumulates sufficient anatomical information to achieve reliable segmentation results, matching the reliability of traditional 3-D methods while maintaining the speed advantage of 2-D processing.
Solution Approach 2:
The patent uses multiple 2-D slices at different positions and orientations to compensate for the information loss in any single 2-D plane. The collective information from multiple 2-D segments provides sufficient context for reliable anatomical structure identification, effectively transforming limited 2-D data into comprehensive 3-D understanding.
3Measurement precision
If 3-D image volume segmentation is used, then segmentation accuracy is improved, but processing time increases
Solution Approach 1:
The patent divides the computationally intensive 3-D volume processing into smaller 2-D slice processing tasks. Each slice is segmented independently and quickly, then the results are assembled into the final 3-D segmentation. This segmentation approach dramatically reduces total processing time while the aggregation of multiple slice results maintains segmentation accuracy.
Solution Approach 2:
The patent performs 2-D segmentation on individual slices before reconstructing the complete 3-D result. By preparing and segmenting slices in advance during the imaging spin process, the system avoids the need for time-consuming post-processing of the entire 3-D volume, thus reducing overall processing time while maintaining accuracy.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
System and method of more efficiently identifying and segmenting anatomical structures from 2-D cone beam CT images, rather than from reconstructed 3-D volume data, is disclosed. An image processing system receives, from a cone beam CT device, at least one 2-D x-ray image, which is part of a set of x-ray images taken from a 360 degree scan of a patient with a cone beam CT imaging device. The x-ray image contains at least one anatomical structure such as vertebral bodies to be segmented. The received x-ray is then analyzed in order to identify and segment the anatomical structure contained in the x-ray image based on a stored model of anatomical structures. Once the 360 degree spin is completed, a 3-D image volume from the x-ray image set is created. The identification and segmentation information derived from the x-ray image is then added to the created 3-D image volume.