3D CT Reconstruction From Two Perpendicular DR Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing three-dimensional CT imaging methods struggle to achieve high-quality imaging quickly due to the need for numerous Digital Radiography (DR) images collected circumferentially, which slows down the reconstruction process.
Innovation Solution
A three-dimensional CT imaging method using two DR images with a perpendicular relationship, combined with a preset volume reconstruction model trained on pseudo-sinograms, to enhance feature extraction and improve reconstruction speed while maintaining image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If numerous DR images are collected circumferentially for three-dimensional CT imaging, then image quality is improved, but reconstruction speed deteriorates
Solution Approach 1:
The patent extracts and utilizes only the essential projection data (two DR images with perpendicular relationships) from the complete circumferential imaging process. By taking out only the necessary minimum data required for reconstruction, the method achieves both high image quality and fast reconstruction speed without needing the full set of circumferential images.
Solution Approach 2:
The patent employs a deep learning-based reconstruction model that learns from training data to generate high-quality three-dimensional volumes from limited projection data. The model creates a computational copy that synthesizes the complete volumetric information from just two perpendicular DR images, effectively replacing the need for actual circumferential imaging.
2Productivity
If two DR images with perpendicular relationship are used for volume reconstruction, then reconstruction speed is improved, but image quality may deteriorate
Solution Approach 1:
The patent incorporates a feedback mechanism through the deep learning model that iteratively refines the reconstruction process. The model receives the two perpendicular DR images as input, processes them through multiple layers of neural networks, and outputs a three-dimensional volume that is continuously optimized to match the expected image quality characteristics, ensuring high quality despite the limited input data.
Solution Approach 2:
The patent introduces a pseudo-sinogram as an intermediary representation that bridges the gap between the limited DR image data and the final three-dimensional volume. This intermediate form allows the reconstruction model to process and interpret the minimal projection data more effectively, transforming it into high-quality volumetric information.
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
This approach allows for rapid generation of high-quality three-dimensional CT imaging by utilizing two DR images with a perpendicular relationship and a trained model, enhancing feature extraction and ensuring efficient reconstruction.
Implementation Method 1
a three-dimensional CT imaging method based on X-ray dual projection
Data Source
AI summary
The disclosure provides a three-dimensional CT imaging method and apparatus. The method includes: collecting two DR images with a perpendicular relationship for an object to be imaged; inputting the two DR images into a preset three-dimensional volume reconstruction model to obtain a three-dimensional volume of the object outputted by the three-dimensional volume reconstruction model; and slicing the three-dimensional volume to obtain three-dimensional CT imaging of the object, wherein the preset three-dimensional volume reconstruction model is trained using multiple training samples, each training sample comprising: a DR image group composed of two DR images with a perpendicular relationship, a pseudo-sinogram corresponding to the DR image group, and a three-dimensional volume corresponding to the DR image group.