X-Ray Scatter Estimation Using Partial Scatter Process Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for scattered radiation correction in X-ray imaging, such as using anti-scatter grids or software-based corrections, either compromise image quality or require high computational overhead, making them unsuitable for low-latency imaging tasks like interventional procedures, and struggle with generalization across different imaging geometries and patient groups.
Innovation Solution
A computer-implemented method using an estimation algorithm that separates partial scattered radiation images for different physical scatter processes, employing machine learning models to determine these images, allowing for efficient and accurate scattered radiation estimation and correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If software-based scattered radiation correction is used, then scattered radiation artifacts are reduced, but computational overhead increases significantly
Solution Approach 1:
The patent uses a trained machine learning model to create a computational copy that replicates the complex physics-based scattered radiation correction. The model learns the relationship between input images and scattered radiation patterns during training, allowing fast inference during actual imaging without requiring real-time complex calculations.
Solution Approach 2:
The patent performs the computationally intensive model training in advance before actual imaging procedures. During training, the model learns from pre-computed ground truth data generated by physics-based simulations or reference measurements. This preliminary action transfers the computational burden from runtime to training time, enabling real-time application during imaging.
2Measurement precision
If machine learning models are trained for specific imaging embodiments, then correction accuracy improves, but generalization to other imaging geometries and patient groups deteriorates
Solution Approach 1:
The patent designs the machine learning model to perform multiple functions across different imaging scenarios. The model is trained on diverse datasets encompassing various imaging geometries, patient groups, and scattering conditions, enabling it to generalize and provide accurate corrections across different application contexts without requiring separate models for each scenario.
Solution Approach 2:
The patent incorporates multiple imaging parameters and geometric configurations into the training data and model architecture. By varying parameters such as source-to-detector distance, detector-to- patient distance, and imaging angles during training, the model learns to handle different imaging embodiments and maintains accuracy across diverse conditions through parameter-aware processing.
Data Source
AI summary
A computer-implemented method for determining an estimation of a two-dimensional scattered radiation distribution in a two-dimensional X-ray image that is based on an imaging procedure performed by an X-ray detector is provided. The method includes receiving the X-ray image, applying an estimation algorithm to the X-ray image or to a two-dimensional intermediate image determined from the X-ray image. The estimation algorithm determines a respective two-dimensional partial scattered radiation image for a plurality of physical scatter processes such that image values of pixels of the respective partial scattered radiation image in each case describe an estimated value for a respective scattered radiation dose that was applied to a respective detector region of the X-ray detector assigned to the respective pixel during acquisition of the X-ray image by the respective physical scatter process. The estimation of the scattered radiation distribution is determined based on the plurality of partial scattered radiation images.