Nonlinear Pixel Enhancement for Multi-Slice CT Spatial Resolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multi-slice CT imaging systems face challenges in improving spatial resolution along the z-axis due to difficulties in designing pre-patient collimators for partial x-ray beam blocking and software-based de-convolution techniques that cause overshoot and undershoot artifacts, leading to potential clinical misinterpretation.
Innovation Solution
A computer-implemented method and system that acquire and reconstruct CT images, identify candidate pixels based on intensity variations, and apply nonlinear enhancement algorithms to iteratively adjust pixel intensities, generating enhanced images without artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If software-based de-convolution techniques are used to reduce slice thickness, then z-axis spatial resolution is improved, but overshoot and undershoot artifacts are introduced
Solution Approach 1:
The patent changes the approach from linear de-convolution to nonlinear iterative enhancement, modifying the mathematical parameters and algorithms used to process the image data. This allows achieving resolution improvement without the harmful artifacts caused by traditional de-convolution methods
Solution Approach 2:
The patent acknowledges that de-convolution inherently causes artifacts but converts this challenge into a benefit by using the artifact patterns themselves as input for iterative enhancement algorithms that learn to distinguish true anatomical features from artifacts, thereby eliminating the harmful effects while preserving the resolution improvement
2Manufacturing precision
If pre-patient collimators are designed to partially block x-ray beams for each detector row, then slice thickness is reduced, but device complexity and manufacturing difficulty increase
Solution Approach 1:
The patent extracts the slice thickness control function from the pre-patient collimator hardware and relocates it to the image processing software stage. This eliminates the need for complex hardware collimators while achieving the same functional outcome through computational methods
Solution Approach 2:
The patent replaces the mechanical pre-patient collimator system with a software-based image processing system. Instead of using physical structures to control beam geometry, the system uses computational algorithms to achieve slice thickness control, thereby reducing device complexity and manufacturing difficulty
3Measurement precision
If slice thickness is reduced to improve spatial resolution, then diagnostic accuracy is enhanced, but noise in the images increases
Solution Approach 1:
The patent performs preliminary actions by collecting redundant projection data from multiple angles and using iterative reconstruction methods before final image generation. This preparatory data collection and processing establishes a foundation that allows noise reduction in the final enhanced image without compromising the improved spatial resolution
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
Improves z-axis spatial resolution in multi-slice CT imaging systems by reducing slice sensitivity profile and noise, resulting in sharper images without overshoot or undershoot artifacts, enhancing diagnostic accuracy.
Implementation Method 1
an x-ray source emits a fan-shaped beam toward a subject or object... The beam, after being attenuated by the subject, impinges upon an array of radiation or x-ray detectors
Implementation Method 2
Each scintillator of a scintillator array converts x-rays to light energy. Each scintillator discharges light energy to a photodiode adjacent thereto
Data Source
AI summary
A system and method include acquisition of a set of projections from an object using a CT imaging system and reconstruct an initial image of the scanned object from the set of projections, the reconstructed initial image comprising a plurality of pixels. The system and method also include identification of a candidate pixel within the plurality of pixels, application of a nonlinear enhancement to the candidate pixel to iteratively adjust an intensity value of the candidate pixel, and generation of a final image using the adjusted intensity value of the candidate pixel.