Respiratory Motion Estimation in Projection Domain Nuclear Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for correcting respiratory motion in nuclear medical imaging, such as SPECT, face challenges due to high noise and low resolution, leading to blurring and artifacts in reconstructed images, and existing motion estimation techniques are inconsistent and unable to adapt to inter-patient variations or changes in respiration during imaging.
Innovation Solution
A sequence-based technique is employed to jointly estimate respiratory motion across multiple projection views using a maximum likelihood objective function and a motion model, incorporating a respiratory surrogate signal to improve data utilization and stability, allowing for patient-specific motion modeling and reduced estimation bias.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If respiratory motion correction is applied using prevailing methods (3D-3D registration or optical flow), then motion artifacts are reduced, but computational runtime increases and consistency deteriorates due to irregular breathing and data deficiencies
Solution Approach 1:
The patent segments the respiratory motion correction process into two distinct stages: (1) motion estimation in the projection domain using 2D-2D registration on individual projection views, and (2) motion correction in the image domain using the estimated motion fields. This segmentation allows each stage to be optimized independently, reducing overall computational complexity while maintaining correction effectiveness.
Solution Approach 2:
The patent transitions from 3D-3D registration (volume-to-volume) to 2D-2D registration (projection-to-projection) for motion estimation. By operating in the projection domain rather than the image domain, the computational complexity is significantly reduced while still capturing the essential motion characteristics needed for correction.
2Object-affected harmful factors
If respiratory motion correction is applied using prevailing methods, then motion artifacts are reduced, but measurement precision deteriorates due to poor quality gated reconstructions from count-deficient or missing data
Solution Approach 1:
The patent performs preliminary motion estimation in the projection domain before conducting image domain correction. By estimating motion fields from projection data (which is less affected by count deficiencies) and then applying these fields to correct the reconstructed images, the method avoids the pitfall of attempting to estimate motion from poor-quality gated reconstructions.
Solution Approach 2:
The patent uses projection domain motion fields as an intermediary to bridge the gap between raw projection data and final image correction. These motion fields serve as a mediator that can be accurately estimated from projection data and then applied to correct images, avoiding direct estimation from problematic image data.
3Object-affected harmful factors
If respiratory motion correction is applied using prevailing methods, then motion artifacts are reduced, but adaptability deteriorates because models are based on averages or assume constant motion between time points
Solution Approach 1:
The patent employs dynamic motion modeling that allows motion parameters to vary across different projection views and time points. Rather than using static average models or assuming constant motion, the method estimates motion fields dynamically from the actual projection data, enabling adaptation to inter-patient variations and changes in respiration patterns during the acquisition.
4Object-affected harmful factors
If respiratory motion correction is applied using prevailing methods, then motion artifacts are reduced, but image quality deteriorates due to high noise in SPECT imaging destabilizing motion estimates
Solution Approach 1:
The patent segments the motion estimation process from the image reconstruction process. By estimating motion fields from projection data (which has higher signal content and is less noisy than reconstructed images) and then applying these stable motion fields to correct the images, the method achieves reliable motion estimation despite the high noise in SPECT imaging.
Data Source
AI summary
In nuclear medical imaging, respiratory motion is corrected. Rather than pairwise estimation of motion from projection views, a sequence-based technique is used to jointly estimate parameters for a motion model across many projection views. In one approach, this sequence-based method is iteratively solved with a maximum likelihood objective function incorporating the motion model. A surrogate respiration signal is used to gate for the joint estimation and used to convert gated motion parameters to temporal motion parameters.