Multi-Resolution Image Reconstruction for CT Quantification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current clinical image acquisition techniques, such as CT scans, face limitations in achieving high resolution for accurate measurement of small objects like coronary plaque and small airways in lungs due to storage and workflow constraints, and existing reconstruction parameters are not suitable for quantitative analysis.
Innovation Solution
A method and system for reconstructing multi-resolution images, where a first three-dimensional image is created at a lower resolution, and specific volumes of interest are reconstructed at a higher resolution using an imaging system with an x-ray source and detector, facilitating quantification of image structures without increasing the number of images required.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If images are acquired at the highest resolution available on a CT system, then measurement precision of small objects is improved, but storage space requirements and workflow complexity increase
Solution Approach 1:
The patent segments the image processing workflow into two distinct stages: a first reconstruction at lower resolution for broad overview and volume identification, and a second reconstruction at higher resolution specifically for identified volumes of interest. This segmentation allows high resolution to be applied selectively rather than universally, reducing overall storage requirements while maintaining measurement precision where needed.
Solution Approach 2:
The patent implements local quality by applying different resolution qualities to different regions of the image data. Volumes of interest undergo second reconstruction at higher resolution, while other regions remain at the lower resolution of the first reconstruction. This ensures that measurement precision is optimized locally for small objects without unnecessarily increasing storage space for entire datasets.
2Measurement precision
If images are acquired at the highest resolution available on a CT system, then measurement precision of small objects is improved, but workflow complexity increases
Solution Approach 1:
The patent performs preliminary action by conducting the first reconstruction at lower resolution first, which enables preliminary identification of volumes of interest before committing to high-resolution reconstruction. This preliminary step guides subsequent high-resolution processing, preventing unnecessary high-resolution reconstruction of regions that do not contain small objects requiring precise measurement.
Solution Approach 2:
The workflow is segmented into distinct processing streams: a rapid low-resolution reconstruction stream for overview and volume identification, and a targeted high-resolution reconstruction stream for measured volumes. This segmentation simplifies workflow management compared to processing entire datasets at high resolution, as each stream has optimized parameters and computational requirements.
3Measurement precision
If a single high resolution reconstruction is performed for the entire volume, then measurement precision is improved, but computing requirements and memory usage increase
Solution Approach 1:
The patent applies partial action by performing high-resolution reconstruction only on identified volumes of interest rather than on the entire imaging volume. The first reconstruction provides sufficient precision for most purposes, and the second high-resolution reconstruction is applied partially only where measurement precision is critically needed, reducing overall computing requirements while maintaining measurement precision for small objects.
Solution Approach 2:
Computational resources are allocated with local quality by directing intensive high-resolution reconstruction computations only to specific volumes of interest containing small objects, while other regions utilize the computationally lighter low-resolution reconstruction. This optimizes the balance between measurement precision and computing requirements by matching computational effort to actual measurement needs.
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 accurate segmentation, classification, and quantification of small plaque deposits and airway measurements by displaying high-resolution areas within a low-resolution image, reducing memory and computing requirements while maintaining overall image context for clinical study.
Implementation Method 1
an x-ray source provided on the rotating member, and an x-ray detector disposed on the rotating member and configured to receive x-rays from the x-ray source
Data Source
AI summary
Methods and apparatus for reconstructing a multiple resolution images of an object are provided. The method includes reconstructing a first three-dimensional image at a first resolution, determining at least one volume of interest in the generated image, and reconstructing a second three-dimensional image of the determined at least one volume of interest at a second resolution, the second resolution being higher than the first resolution such that a quantification of image structures is facilitated.