Partial Volume Correction in PET Penalized-Likelihood Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Partial volume errors (PVEs) in Positron Emission Tomography (PET) image reconstruction are challenging to correct due to their complex dependence on patient size, photon count, ROI location, background activity, and reconstruction parameters, especially in iterative image reconstruction methods which are nonlinear and space-variant.
Innovation Solution
A computer-implemented method for partial volume correction using a pre-calculated contrast recovery coefficient (CRC) value, stored in a lookup table (LUT), which corrects quantitation by accounting for PVEs in PET image reconstruction. The CRC value is calculated based on simulated lesions with varying sizes and activity concentrations, and used to correct uncorrected quantitation in reconstructed PET images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If standard PET image reconstruction methods are used, then image reconstruction is achieved, but partial volume errors occur due to finite spatial resolution
Solution Approach 1:
The patent pre-calculates contrast recovery coefficients for various lesion sizes and stores them in lookup tables before actual image reconstruction. This preliminary computation allows the complex PVC correction to be performed efficiently during quantitation without requiring complex real-time calculations, thus improving quantitation accuracy while managing computational complexity
Solution Approach 2:
The patent introduces contrast recovery coefficients as intermediary values that mediate between the true activity concentration and the reconstructed image values. These coefficients act as correction factors that translate the relationship between ground truth and reconstructed images into a simple multiplicative or additive correction, simplifying the overall correction process while maintaining accuracy
2Reliability
If iterative image reconstruction methods are used, then image quality improves, but partial volume errors become more complex and difficult to correct
Solution Approach 1:
The patent performs preliminary calculations of contrast recovery coefficients using simulated data with known ground truth before actual clinical imaging. By pre-computing and storing these coefficients in lookup tables indexed by lesion size and reconstruction parameters, the system can quickly apply corrections during actual imaging without re-running complex iterative simulations, thus maintaining high image quality while simplifying the correction process
Solution Approach 2:
The patent creates simulated phantom images with known activity distributions to copy and study the behavior of iterative reconstruction algorithms under controlled conditions. These simulations allow the system to pre-determine contrast recovery coefficients that can be applied to actual patient data, separating the complex simulation work from the clinical application and enabling efficient correction
3Measurement precision
If partial volume correction is applied, then quantitation accuracy improves, but computation time increases
Solution Approach 1:
The patent performs all complex computational work in advance by pre-calculating contrast recovery coefficients for various lesion sizes and storing them in lookup tables. During actual image processing, the system only needs to retrieve pre-computed coefficients and apply simple corrections, dramatically reducing computation time while maintaining high quantitation accuracy
Solution Approach 2:
The patent pre-calculates contrast recovery coefficients for a comprehensive range of lesion sizes and conditions that may exceed the actual clinical needs. By storing these pre-computed values in lookup tables, the system can quickly retrieve appropriate coefficients for specific cases without performing full recalculations, thus reducing computation time while ensuring accuracy across different scenarios
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 method achieves accurate and consistent quantitation by systematically correcting for PVEs, improving the reliability of PET image analysis, particularly in iterative image reconstruction methods.
Implementation Method 1
A radionuclide in the patient's body decays and emits a positron
Implementation Method 2
which undergoes an annihilation with a nearby electron, subsequently generating a pair of 511 keV gamma ray photons
Implementation Method 3
If a pair of gamma ray photons are detected within a coincidence timing window by two in an array of PET detectors, a coincidence event is recorded
Implementation Method 4
reconstructing the activity distribution from the emission data by maximizing a penalized-likelihood objective function to produce a reconstructed PET image
Data Source
AI summary
A computer-implemented method for partial volume correction in Positron Emission Tomography (PET) image reconstruction includes receiving emission data related to an activity distribution, reconstructing the activity distribution from the emission data by maximizing a penalized-likelihood objective function to produce a reconstructed PET image, quantifying an activity concentration in a region of interest of the reconstructed PET image to produce an uncorrected quantitation, and correcting the uncorrected quantitation based on a pre-calculated contrast recovery coefficient value to account for a partial volume error in the uncorrected quantitation.