Image-Based Attenuation Correction for PET/SPECT Misalignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current PET and SPECT imaging systems face challenges in accurately correcting for attenuation mismatch between CT and PET images due to patient movement and misalignment, leading to artifacts and unreliable data analysis, especially when raw emission data is not readily available.
Innovation Solution
A system and method for image-based correction that allows for offline processing of PET and SPECT images using existing data storage systems, where processors determine and calculate attenuation mismatch estimates to generate corrected images without requiring access to raw emission data, facilitating easier incorporation into clinical workflows and improving registration techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If attenuation correction is performed using CT images in PET-CT scanners, then structural details and functional features can be precisely located, but patient movement and misalignment between CT and PET scans cause attenuation mismatch leading to artifacts and unreliable data analysis
Solution Approach 1:
The patent implements dynamic attenuation correction by registering the attenuation image to the emission image using motion estimation techniques. This allows the attenuation correction to adapt to patient movement during the scan, dynamically adjusting the alignment between CT and PET data to maintain both precision in locating features and reliability in data analysis throughout the scanning process
Solution Approach 2:
The system uses feedback from motion detection algorithms to identify misalignment between CT and PET images, then applies corrective transformations to the attenuation image. This feedback loop continuously monitors and corrects for patient movement, ensuring that attenuation correction remains accurate despite physiological motion during scanning
2Measurement precision
If re-reconstruction of emission images is performed to correct attenuation mismatch, then image accuracy can be improved, but access to raw emission data is required which is not always available and increases processing complexity
Solution Approach 1:
The patent extracts and corrects attenuation mismatch errors directly from the reconstructed emission images without requiring access to raw emission data. By working with the already-reconstructed images and applying correction factors derived from motion estimation, the system achieves improved accuracy while avoiding the complexity of re-reconstruction pipelines and raw data access requirements
3Adaptability or versatility
If CT and PET scans are performed sequentially, then both structural and functional imaging can be obtained, but patient movement between scans causes misalignment and attenuation errors
Solution Approach 1:
The patent performs preliminary motion estimation and attenuation image registration before final image fusion. By pre-aligning the attenuation image to the emission image using motion detection algorithms, the system compensates for patient movement that occurs during the sequential scanning process, maintaining high alignment precision while preserving the versatility of combined structural and functional imaging
Data Source
AI summary
A system and method for image-based correction including a receiver to acquire an image from one or more data storage systems, one or more processors to determine an attenuation mismatch estimate and calculate a correction for the image based on the attenuation mismatch estimate and the image, and an output to generate an attenuation mismatch corrected image based on the correction.