X-Ray Marker Localization in 3D Space with Motion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for identifying marker points in x-ray images are slow and inaccurate due to the non-linear nature of x-ray imaging and large dynamic ranges, leading to positional errors and blurring in images, especially when markers are close together or subject to patient movement.
Innovation Solution
The proposed solution involves a method that reduces image processing time by applying a cumulative mean filter and high-pass filtering to isolate circular marker points, followed by sub-pixel localization using marker point profiles and cylindrical coordinate fitting to accurately determine marker positions, even in the presence of noise and movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional methods are used to identify marker points in x-ray images, then marker localization is performed, but processing time is excessive and accuracy deteriorates due to non-linear imaging characteristics and large dynamic ranges
Solution Approach 1:
The patent segments the complex marker identification problem into distinct processing stages: background modeling, marker candidate detection, and precise localization. Each stage handles specific aspects of the problem independently, improving both speed and accuracy by avoiding monolithic processing of the entire image with complex algorithms.
Solution Approach 2:
The patent transforms the non-linear x-ray image data into a linearized representation by applying logarithmic transformation and background subtraction. This parameter change converts the difficult non-linear localization problem into a simpler linear problem that can be solved quickly and accurately using standard techniques.
2Measurement precision
If traditional marker identification algorithms are applied, then marker positions are detected, but accuracy deteriorates when markers are close together or subject to patient movement
Solution Approach 1:
The patent performs preliminary background modeling and subtraction before marker detection. By removing the background structure in advance, the algorithm creates a cleaner signal that makes subsequent marker detection more reliable, even when markers are close together or when patient movement occurs during imaging.
Solution Approach 2:
The patent implements an iterative refinement process where initial marker detections are used to update the background model, which then improves subsequent detections. This feedback loop continuously enhances accuracy by incorporating information from previous detection iterations, making the system more robust to motion and closely-spaced markers.
3Measurement precision
If comprehensive image processing is performed to handle complex backgrounds and noise, then marker localization accuracy improves, but processing complexity and time increase
Solution Approach 1:
The patent extracts and removes the background component from the x-ray image before performing marker localization. By separating the background structure from the marker signals and eliminating it through subtraction, the system simplifies the processing task while maintaining high accuracy in complex backgrounds, avoiding the need for extremely complex full-image analysis algorithms.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method and system of determining a radial distance (R), an angular position (f), and a axial position (Z) of a marker identified in a sequence of projection images. Embodiments of the invention allow a marker three-dimensional localization module executable by the electronic processing unit to obtain a sequence of images based on image data generated by a scanner. Each image in the sequence of images represents an angle of rotation by the scanner and includes a marker point position. The behavior of first values and second values of the marker point positions are analyzed through the sequence of images to determine the radial distance of the marker, the angular position of the marker, and the axial position of the marker. Thus, embodiments of the invention allow for rapidly detecting and localizing external markers placed on a patient in projection images.