Energy-Based Scatter Fraction Estimation for PET Image Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing scatter correction algorithms in PET scans are inaccurate due to neglect of multiple-scattering and improper scaling of the estimated scatter sinogram, leading to degraded image quality and quantitative accuracy.
Innovation Solution
A method for estimating the singles scatter fraction using list-mode data from PET scans, which involves determining the number of events in specific energy windows and calculating the singles scatter fraction to properly scale the scatter sinogram, allowing for accurate scatter correction during the reconstruction process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If analytic-model-based scatter correction algorithm is used, then calculation speed is improved, but accuracy deteriorates due to improper scaling of estimated scatter sinogram
Solution Approach 1:
The patent changes the parameter of scatter fraction estimation from conventional analytic models to energy-based methods using list-mode data. By calculating scatter fraction from the ratio of events in different energy windows (Compton window vs. photopeak window), the system achieves both speed and accuracy improvement, resolving the contradiction between calculation speed and accuracy.
2Speed
If multiple-scattering is neglected to improve calculation speed, then calculation speed is improved, but accuracy deteriorates due to neglect of multiple-scattering events
Solution Approach 1:
The patent employs a self-service approach where the system uses its own detected events (singles events in Compton and photopeak windows) to automatically estimate the scatter fraction. This self-estimation mechanism eliminates the need for complex external modeling or simulation, achieving both speed and accuracy by leveraging the scanner's own data without requiring additional computational resources or assumptions about multiple-scattering patterns.
3Ease of operation
If energy-based scatter correction algorithm is used, then practicality is improved, but accuracy deteriorates compared to analytic-model-based methods
Solution Approach 1:
The patent replaces the conventional analytic-model-based mechanical calculation system with an energy-based statistical approach. Instead of using complex mathematical models with scaling factors, the system substitutes a direct statistical estimation method using list-mode data and energy window ratios. This substitution achieves both practicality and accuracy by using the scanner's inherent energy discrimination capability to directly estimate scatter fractions without relying on approximate physical models.
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 improves the accuracy of scatter correction, enhancing the quality and quantitation of PET images by accurately estimating the scatter fraction for each crystal or group of crystals, thereby improving image reconstruction.
Implementation Method 1
Compton scattering in a PET scan of a patient degrades the image quality of a reconstructed PET image
Implementation Method 2
a first number of events occurring in a first energy window spanning a first energy range
Data Source
AI summary
A method is provided for determining a scatter fraction for a radiation diagnosis apparatus. The method includes acquiring an energy spectrum from list mode data obtained from a scan performed using the radiation diagnosis apparatus; determining, from the acquired list mode data, a first number of events occurring in a first energy window spanning a first energy range; determining, from the acquired list mode data, a second number of events occurring in a second window, the second energy window spanning a second energy range different from the first energy range; calculating a singles scatter fraction based on the determined first number of events and the determined second number of events; and reconstructing an image based on the acquired list mode data and the calculated singles scatter fraction.