Imaging ROI Segmentation for Automatic Monitoring Slice Positioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current contrast enhancement scanning methods rely heavily on manual setting of monitoring layers and regions of interest, which are subjective to the technician's experience and increase workload, affecting image quality.
Innovation Solution
An imaging system and method that uses a neural network to automatically determine the monitoring layer and region of interest through cascaded coarse and fine segmentation, employing techniques like threshold, edge detection, and deep learning to accurately segment the target region of interest.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual setting of monitoring layer and region of interest is used, then image quality can be optimized, but technician workload increases and accuracy depends on experience
Solution Approach 1:
The system performs self-service by automatically detecting the monitoring layer and region of interest through neural network processing of positioning images, eliminating the need for manual technician intervention. The neural network autonomously identifies target regions and determines optimal monitoring parameters based on the positioning image data.
Solution Approach 2:
The patent replaces the mechanical/manual system with an automated neural network-based system. Instead of manual selection by technician, the neural network processes positioning images to automatically determine monitoring layers and regions of interest, substituting human expertise with an automated intelligent system.
2Reliability
If manual setting of monitoring layer and region of interest is used, then image quality can be optimized, but reliability depends on technician experience
Solution Approach 1:
The system achieves self-service automation where the neural network independently processes positioning images to determine monitoring parameters without relying on technician experience. This ensures consistent and reliable results that do not vary based on individual operator skill levels.
Solution Approach 2:
The system incorporates feedback mechanisms where the neural network processes positioning images, determines monitoring layers and regions of interest, and these parameters are then used to guide the contrast enhancement scanning process. This feedback loop ensures reliable and consistent image quality through automated parameter optimization.
3Productivity
If automated neural network segmentation is used, then technician workload is reduced, but system complexity increases
Solution Approach 1:
The patent applies segmentation by dividing the image processing task into distinct stages: positioning image acquisition, neural network-based monitoring layer determination, region of interest identification, and contrast enhancement scanning. This segmentation of the overall process into manageable modules reduces the complexity burden while maintaining high productivity.
Solution Approach 2:
The system performs preliminary action by using positioning images to pre-determine monitoring layers and regions of interest before the actual contrast enhancement scanning begins. This preliminary automated setup reduces the complexity of the main scanning process and improves overall processing efficiency.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure relates to an imaging system and method. Specifically, an imaging system comprises: a positioning image acquisition unit, configured to acquire a positioning image of a scanning object; a monitoring slice image acquisition unit, configured to determine a key point corresponding to the position of a target region of interest in the positioning image by using a neural network, and acquire a monitoring slice image of the scanning object at the position of the key point; and a target region-of-interest segmentation unit, configured to segment the monitoring slice image to obtain the target region of interest. The present invention can accurately acquire the position of the monitoring slice, and can accurately obtain the target region of interest through segmentation by a cascaded coarse segmentation and fine segmentation.