TOF Scatter Distribution Estimation in PET Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current time-of-flight (TOF) positron emission tomography (PET) systems face challenges in efficiently estimating scatter distribution due to resource-intensive single scatter simulations, which are computationally demanding and produce noise, especially in low activity regions and low count data.
Innovation Solution
The technology employs non-TOF projection data reduction and unbiased reconstruction using NEG-OSEM or FBP+FORE algorithms to estimate TOF scatter distribution, followed by smoothing filtering to reduce noise, avoiding the need for TOF modeling and improving scatter correction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If single scatter simulation is used for TOF scatter distribution estimation, then scatter correction accuracy is improved, but computational resource consumption increases and noise is produced in low activity regions
Solution Approach 1:
The patent segments the scatter estimation process into non-TOF projection data reduction, unbiased reconstruction using NEG-OSEM or FBP+FORE algorithms, and subsequent smoothing filtering. This segmentation allows the system to avoid computationally intensive TOF modeling while maintaining accuracy through targeted processing stages.
Solution Approach 2:
The patent creates a simplified copy of the scatter estimation process by using non-TOF projection data instead of full TOF data. This copying approach preserves the essential scatter information needed for correction while eliminating the computational burden of TOF modeling, effectively resolving the contradiction between accuracy and resource consumption.
2Measurement precision
If single scatter simulation is used for TOF scatter distribution estimation, then scatter correction accuracy is improved, but noise is produced in low activity regions and low count data
Solution Approach 1:
The patent performs preliminary smoothing filtering on the scatter distribution estimated from non-TOF projection data before using it for correction. This preliminary action reduces noise in low activity regions while preserving the accurate scatter correction benefits, effectively eliminating the harmful noise effect.
3Productivity
If non-TOF projection data reduction and unbiased reconstruction are used, then computational resources are reduced and noise is reduced, but TOF modeling is avoided which may affect accuracy
Solution Approach 1:
The patent changes the parameter space by working with non-TOF projection data instead of TOF data, and uses unbiased reconstruction algorithms (NEG-OSEM or FBP+FORE) to maintain accuracy. This parameter change allows the system to achieve both computational efficiency and accurate scatter estimation without requiring TOF modeling.
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 results in accurate scatter estimation with reduced noise and computational resources, achieving image quality comparable to standard TOF processing, even in low count data scenarios, and effectively corrects for scatter inconsistencies.
Implementation Method 1
Time of flight scatter distribution estimation in positron emission tomography
Data Source
AI summary
Estimating time-of-flight (TOF) scatter distribution in a positron emission tomography (PET) system. Obtaining PET TOF projection data: PET random coincidence data and PET TOF prompt coincidence events data. Reducing measured TOF projection data to non-TOF projection data. Reconstructing, unbiased, the non-TOF projection date. Forward projecting unbiased reconstructed non-TOF projection data to estimate TOF trues distribution. Subtracting: The estimated TOF trues distribution and the measured random coincidence, from measured TOF prompt coincidence events.