Adaptive Tomogram Scanning via Entropy-Based ROI Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-resolution scanning of large regions becomes burdensome as resolution increases, often revealing that lower resolutions could suffice in certain sub-regions, leading to inefficiencies in imaging and reconstruction processes.
Innovation Solution
A method dynamically determines where high- and low-resolution scans are necessary during imaging by identifying regions of interest based on entropy and gradient thresholds, allowing for iterative adjustments in resolution to optimize scanning efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-resolution scanning is performed across the entire large region, then measurement precision is improved, but use of energy increases and productivity decreases
Solution Approach 1:
The patent applies local quality by performing high-resolution scanning only in identified regions of interest rather than uniformly across the entire large region. The system first performs a low-resolution scan to identify areas requiring detailed examination, then concentrates high-resolution scanning resources only on those specific sub-regions where entropy and gradient thresholds indicate potential defects or features of interest.
2Measurement precision
If high-resolution scanning is performed across the entire large region, then measurement precision is improved, but productivity decreases
Solution Approach 1:
The system implements local quality by applying different resolution levels to different spatial regions. High-resolution scanning is selectively applied only to identified regions of interest, while the remainder of the large region is scanned at lower resolution, thereby maintaining measurement precision where needed while significantly improving overall scanning productivity.
Solution Approach 2:
The patent employs preliminary action by first performing a low-resolution preliminary scan of the entire large region to identify areas of interest based on entropy and gradient analysis. This preliminary action guides subsequent high-resolution scanning, ensuring that detailed examination is focused only on regions where it is truly necessary, thus optimizing productivity.
3Measurement precision
If high-resolution scanning is performed across the entire large region, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The system applies dynamics by making the scan resolution adaptive rather than static. The resolution level is dynamically adjusted based on the content of the region being scanned, with high resolution applied only to regions of interest identified through real-time entropy and gradient analysis, and lower resolution used for the remainder of the large region.
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 reduces the burden of high-resolution scanning across large areas by focusing higher resolution on specific sub-regions, enabling more efficient scanning with lower power devices and resulting in a more compact and portable scanning unit.
Implementation Method 1
acquiring a plurality of first x-ray attenuation images of a subject, the plurality of first x-ray attenuation images suitable for reconstructing a first density function indicative of attenuation of the x-ray radiation
Data Source
AI summary
The present invention seeks to reduce the burden of producing high-resolution tomograms by using an initial scan on a predetermined grid 10 to obtain a minimal set of images, and then regions of interest 20 are identified for further scanning. The further scanning locations 40 are determined by image entropy or gradient found in the previous iteration; such regions are indicative of edges, cracks or complex structure within the region. After each iteration, the level of information (e.g. image entropy or gradient) will decrease relative to the pixel/voxel size. In this way, a more efficient way to scan is achieved.