Nuclear Medical Image Reconstruction With Inter-Step Motion Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Nuclear medical imaging methods face challenges in maintaining image quality and spatial correspondence due to patient motion between different acquisition steps, especially in dynamic and combined-modality workflows, leading to blurring effects and quantification biases in final summed images.
Innovation Solution
A computer-implemented method that utilizes motion information from different acquisition steps to correct for patient motion, employing motion-sensitive iterative reconstruction techniques and combining data from multiple passes to generate high-quality, spatially consistent nuclear medical image data sets, leveraging integrated PET-MR devices for additional motion and attenuation correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If nuclear medical raw data are acquired in multiple acquisition steps separated in time to cover larger regions of interest, then the coverage area increases, but patient motion occurs between acquisition steps causing image quality degradation
Solution Approach 1:
Motion data is determined from source data before the final image reconstruction process. By preliminarily analyzing the motion between acquisition steps using source data (such as attenuation maps or images from the additional modality), the system prepares motion correction information in advance that can be applied during reconstruction to prevent motion-induced image quality degradation.
Solution Approach 2:
Motion data acts as an intermediary between the multiple acquisition steps and the final image reconstruction. This motion information, derived from source data, mediates the correction process by providing transformation information that aligns data from different time points, thereby resolving the conflict between extended coverage and image quality.
2Loss of time
If attenuation maps are acquired at the beginning of the examination in a first acquisition step, then the acquisition time is reduced, but the attenuation maps may not represent the motion state of later acquisition steps causing quantification biases
Solution Approach 1:
Motion data serves as an intermediary to update the attenuation correction information. Instead of acquiring new attenuation maps at each acquisition step, the system uses motion data derived from source data to transform and update the initial attenuation map, making it representative of the patient's actual position and anatomy at later time points without additional acquisition time.
Solution Approach 2:
The attenuation map parameters are dynamically adjusted based on motion data. By applying motion transformations to the attenuation map, the system changes the spatial parameters of the attenuation correction data to match the patient's position at different acquisition steps, thereby maintaining quantification accuracy without extending acquisition time.
3Manufacturing precision
If motion correction is applied within acquisition steps for physiological motions, then intra-acquisition image quality improves, but inter-acquisition motion between different acquisition steps remains uncorrected
Solution Approach 1:
The invention extends motion correction from the temporal dimension (within a single acquisition step) to the inter-acquisition dimension. By determining motion data between different acquisition steps using source data, the system adds a new dimension of correction that addresses spatial misalignment between separately acquired datasets, complementing the existing intra-acquisition motion correction.
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
Computer-implemented method for determining at least one nuclear medical image data set (6) in nuclear medical imaging using an imaging device (14), wherein - multiple nuclear medical raw data sets (9) of a region of interest of a patient (21) are acquired in respective acquisition steps (2) during the progression of a tracer in the region of interest, - at least one nuclear medical image data set (6) is reconstructed from the nuclear medical raw data sets (9) for each acquisition step (2), - motion data (8) describing the motion of the patient (21) between acquisition steps (2) is determined from source data (12) describing the motion in the region of interest and motion correction is applied to the series of nuclear medical raw data sets (9) and/or the at least one nuclear medical image data set (6) according to the motion data (8).