Dynamic PET Frame Clustering for Reconstruction Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional dynamic frame reconstruction in PET imaging is computationally expensive due to the need for scatter and random estimations for each frame, leading to long processing times for generating medical images.
Innovation Solution
Frames with similar reconstruction parameters are grouped using clustering techniques, allowing scatter and random estimations to be computed once and shared among frames in each group, reducing computational burden while maintaining image accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If scatter and random estimations are computed for each dynamic frame, then image reconstruction accuracy is maintained, but computation time increases significantly
Solution Approach 1:
The patent segments the dynamic frames into groups based on similarity of reconstruction parameters. Frames with similar characteristics are clustered together, allowing scatter and random estimations to be computed once per group rather than for each individual frame. This segmentation approach maintains accuracy within groups while significantly reducing overall computation time.
Solution Approach 2:
The patent merges the scatter and random estimation computations for multiple frames into a single computation per frame group. By combining frames with similar reconstruction parameters into groups, the system performs one estimation per group that applies to all frames in that group, thereby reducing redundant computations while preserving reconstruction accuracy.
2Manufacturing precision
If dynamic frames are reconstructed individually with full scatter correction, then image quality is maintained, but processing speed decreases
Solution Approach 1:
The patent divides the sequence of dynamic frames into groups based on temporal and parameter similarity. This segmentation allows the system to process frames in batches rather than individually, maintaining image quality through group-specific scatter correction while improving processing speed by reducing the number of separate reconstruction operations.
Solution Approach 2:
The patent performs preliminary clustering of frames into groups before executing the full reconstruction process. This preliminary organization allows scatter and random estimations to be pre-computed for each group, which then can be efficiently applied to all frames in the group during reconstruction, thereby maintaining image quality while accelerating processing.
3Measurement precision
If scatter correction is performed for each frame, then reconstruction accuracy is maintained, but computational complexity increases
Solution Approach 1:
The patent merges the computational complexity of scatter correction by performing a single estimation per frame group rather than for each individual frame. This merging approach reduces the total number of complex computations required while maintaining reconstruction accuracy within each group, thereby reducing overall computational complexity.
Solution Approach 2:
The patent changes the parameter approach by using group-specific scatter and random estimation parameters rather than frame-specific parameters. This parameter change from individual frame level to group level maintains the essential reconstruction accuracy while significantly reducing the number of parameters that need to be computed and managed, thereby reducing computational complexity.
Data Source
AI summary
A dynamic frame reconstruction apparatus and method for medical image processing is disclosed which reduces the computationally expensive reconstruction of images but which retains the accuracy of the image reconstruction. A convolutional neural network is used to cluster the dynamic data into groups of frames, each group sharing similar radiotracer distribution. In one embodiment, groups of frames that have similar reconstruction parameters are determined, and scatter and random estimations are computed once and shared among each of the frames in the same frame group.