PET/MR Image Registration with Dynamic Reference Updates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing PET/MR imaging methods struggle with motion artifacts due to patient movement, particularly in dynamic PET imaging where tracer distribution changes, leading to inaccurate image registration and quality degradation.
Innovation Solution
A method for dynamic PET/MR imaging that updates the reference frame throughout the scan by incorporating recently co-registered images, prioritizing dominant motion parameters and dividing the field of view into sections for individual motion correction, using PET data for real-time motion estimation and MR sequence updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single static reference frame is used for co-registration, then the device complexity is reduced, but the manufacturing precision (registration accuracy) deteriorates when tracer distribution changes during acquisition
Solution Approach 1:
The patent applies the dynamics principle by transitioning from a static reference frame to a dynamic reference frame that is continuously updated throughout the PET acquisition. The reference frame is regenerated at predetermined time points using recently co-registered frames, allowing it to adapt to changing tracer distribution. This dynamic update mechanism resolves the contradiction by introducing necessary complexity to maintain high registration accuracy when tracer distribution evolves during the scan.
Solution Approach 2:
The patent implements feedback by using the most recently co-registered frames to update the reference frame. This feedback loop ensures that the reference frame continuously reflects the current tracer distribution, improving registration accuracy. The system monitors co-registration quality and uses this information to regenerate the reference frame at appropriate intervals, balancing the need for accuracy with computational complexity.
2Manufacturing precision
If the reference frame is dynamically updated throughout the scan, then the manufacturing precision (registration accuracy) is improved, but the device complexity increases
Solution Approach 1:
The patent applies periodic action by updating the reference frame at predetermined time points during the PET acquisition rather than continuously. This periodic update strategy maintains high registration accuracy while reducing computational complexity compared to continuous updates. The system balances the frequency of reference frame regeneration to match the dynamics of tracer distribution changes.
Solution Approach 2:
The patent uses partial action by incorporating only the most recently co-registered frames (e.g., 1-3 frames) to update the reference frame, rather than using all available data. This partial update approach maintains registration accuracy while reducing the computational burden of processing the entire dataset, thus managing device complexity effectively.
3Measurement precision
If optical camera systems are used for real-time motion tracking, then the measurement precision of head motion is improved, but the PET signal quality deteriorates due to camera attenuation
Solution Approach 1:
The patent extracts the motion tracking function from the optical camera system and implements it using PET data itself. By deriving motion information from the PET frames through co-registration analysis, the system eliminates the need for external camera systems that cause signal attenuation. This extraction of the motion tracking function resolves the contradiction by removing the harmful attenuation effect while maintaining motion measurement capability.
Solution Approach 2:
The patent uses PET data as an intermediary to obtain motion information without requiring direct optical measurement. The co-registration of PET frames serves as the intermediary process that indirectly measures head motion while avoiding the harmful effects of camera systems. This intermediary approach maintains measurement precision while eliminating signal attenuation.
Data Source
AI summary
A method for combined PET/MR imaging performs a PET/MR scan using a PET/MR scanner. During the scan a frame is reconstructed and co-registered with a reference frame. The reference frame is dynamically updated throughout the PET/MR scan by incorporating most recently co-registered images from previously reconstructed frames.