Cone Beam CT Projection Matrix Update via DRR Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In cone beam CT systems, accurately determining the positional relationship among the X-ray source, detector, and subject is challenging due to structural deflection or deformation of the gantry, leading to reduced quality of three-dimensional images.
Innovation Solution
An imaging system that generates and updates a projection matrix using light source and detector positions, revolution angles, and pixel values to improve the accuracy of three-dimensional image generation, incorporating modules for projection matrix generation, volume image creation, and digitally reconstructed radiograph matching to enhance image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the gantry rotates along a predetermined track to acquire projection images, then the imaging process can be completed, but structural deflection or deformation of the gantry causes inaccurate determination of the rotary orbit, leading to reduced quality of three-dimensional images
Solution Approach 1:
The patent applies feedback by generating a digitally reconstructed radiograph (DRR) from the initial volume image and comparing it with the actual projection image. The difference between the DRR and projection image is used to update and refine the projection matrix, creating a closed-loop feedback mechanism that continuously improves the accuracy of positional relationship determination and thereby enhances three-dimensional image quality despite gantry deformation
Solution Approach 2:
The patent creates a digital copy (DRR) of the projection image from the reconstructed volume image. This digital replica is then compared with the actual projection image to identify discrepancies caused by gantry deformation. By using this copied representation, the system can iteratively refine the projection matrix without requiring perfect mechanical precision of the gantry
2Productivity
If the positional relationship among the X-ray source, detector, and subject is not accurately determined due to gantry deformation, then the imaging process can still proceed, but the quality of the three-dimensional image is reduced
Solution Approach 1:
The patent replaces reliance on mechanical precision (gantry track accuracy) with computational methods. Instead of depending on the mechanical system to maintain exact positional relationships, the system uses mathematical algorithms to generate projection matrices and DRRs, comparing and refining these computationally to achieve accurate three-dimensional reconstruction even when mechanical positioning is imperfect
3Loss of time
If a projection matrix is generated based on assumed positional relationships, then volume images can be generated quickly, but inaccurate positional data leads to reduced image quality
Solution Approach 1:
The patent performs preliminary action by initially generating a volume image and DRR using an assumed projection matrix based on predetermined positional relationships. This initial reconstruction serves as a starting point that can be quickly generated, and then refined through iterative comparison and matrix updating, allowing the system to proceed without waiting for perfect positional measurement while still achieving high-quality results
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system effectively improves the quality of three-dimensional images by accurately accounting for positional relationships, even when gantry orbits are inaccurately determined, through iterative updates of the projection matrix, resulting in clearer and more precise volume images.
Implementation Method 1
illuminating, by a light illuminator, light onto a subject from a light source
Data Source
AI summary
Provided is an operating method of an imaging system, the operating method including illuminating, by a light illuminator, light onto a subject from a light source, acquiring, by an optical detector, a two-dimensional projection image on the subject, generating a first projection matrix between three-dimensional coordinates of the subject and two-dimensional coordinates of the projection image, generating a three-dimensional first volume image on the subject on the basis of the first projection matrix and the projection image, generating a two-dimensional digitally reconstructed radiograph (DRR) on the subject from the first volume image, matching the projection image and the DRR, and updating the first projection matrix on the basis of a matched result to generate a second projection matrix, and generating a three-dimensional second volume image on the subject on the basis of the second projection matrix and the projection image.