Parametric PET Reconstruction with Continuous Bed Motion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Continuous bed motion (CBM) PET systems face challenges in accurate image reconstruction due to non-uniform time information across axial slices, leading to image non-uniformity and incorrect quantification, which complicates parametric imaging and tumor detection.
Innovation Solution
A method that records finely sampled bed tags to accurately track position and time information, calculating slice acquisition times based on average tracer activity, and applies the Patlak model to reconstruct parametric images for metabolism rate and distribution volume, ensuring consistent quantification across different scan modes and organs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If continuous bed motion is used to acquire whole body images, then the field of view is extended to cover the entire patient body, but the time information becomes non-uniform across axial slices leading to image non-uniformity and incorrect quantification
Solution Approach 1:
The patent segments the continuous bed motion scan into multiple axial slices, each with its own acquisition time calculation. By treating each slice independently with slice-specific timing information, the system maintains quantification accuracy across the entire field of view while accounting for the continuous motion of the bed.
Solution Approach 2:
The patent introduces a time dimension to the spatial reconstruction process by calculating slice acquisition times based on bed position and scan velocity. This temporal information is integrated into the reconstruction algorithm, transforming the problem from a purely spatial reconstruction to a spatio-temporal reconstruction that accounts for continuous bed motion.
2Adaptability or versatility
If continuous bed motion with varying velocity is used, then flexible scan modes are achieved, but the position intervals between successive measurements become non-uniform complicating the reconstruction process
Solution Approach 1:
The patent changes the parameter representation from uniform spatial sampling to non-uniform temporal sampling. By recording bed position and time for each measurement and using these parameters to calculate slice acquisition times, the system accommodates variable scan velocities and flexible scan modes while maintaining reconstruction accuracy through parameter-based correction.
3Device complexity
If traditional SUV imaging is used, then the imaging process is simple, but parametric imaging capabilities and tumor detection accuracy are limited
Solution Approach 1:
The patent performs preliminary calculation of slice acquisition times and integration of blood input function values at these specific time points before the actual image reconstruction. This preliminary temporal correction is built into the reconstruction pipeline, allowing parametric imaging to be performed with accuracy comparable to or better than traditional SUV imaging while maintaining a streamlined workflow.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
A method of processing and reconstructing dynamic positron emission tomography (PET) sinogram data comprises: acquiring PET sinogram data using continuous bed motion having a varying velocity; recording a plurality of position-time coordinate pairs while acquiring the PET sinogram data; determining respective acquisition times of each of a plurality of slices of the image, based on the plurality of position-time coordinates; and reconstructing respective parametric images for each respective slice in the plurality of slices.