Spatio-temporal Image Reconstruction via Sparse Regression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In medical imaging, there is a tradeoff between spatial and temporal resolution, where improving one dimension often compromises the other, and there is a need to reduce scan time to minimize patient exposure to harmful radiation.
Innovation Solution
The technique reconstructs spatio-temporal images by using captured data from multiple frames across different time intervals, incorporating both spatial and temporal redundancy, and leveraging secondary information to improve image quality, allowing for enhanced resolution without increasing scan time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If the number of frames captured within a given time period is increased to improve temporal resolution, then temporal resolution is improved, but the amount of data captured in each frame decreases, leading to poorer spatial resolution
Solution Approach 1:
The patent merges data from multiple frames captured at different time intervals to reconstruct a single spatial image. By combining the data from these multiple frames, the system achieves high temporal resolution (using many frames over time) while maintaining high spatial resolution (through the combined data from all frames). This directly resolves the contradiction by allowing both temporal and spatial resolution to be improved simultaneously rather than forcing a tradeoff.
2Object-affected harmful factors
If the scan time is reduced to minimize patient exposure to harmful radiation, then radiation exposure is reduced, but the quality and completeness of captured data may deteriorate
Solution Approach 1:
The patent uses secondary information from external sources to pre-establish constraints and guidance for the image reconstruction process. This preliminary information allows the system to accurately reconstruct images even when the captured data is incomplete due to reduced scan time, thereby maintaining data quality while minimizing radiation exposure.
Solution Approach 2:
The patent introduces secondary information from external sources as an intermediary to bridge the gap between reduced scan time and maintained image quality. This external information acts as a mediator that provides additional constraints and guidance, allowing accurate reconstruction even with limited captured data from the shortened scan.
3Ease of manufacture
If spatial images are reconstructed individually on a frame-by-frame basis, then the reconstruction process is simple and straightforward, but both spatial and temporal redundancy are not exploited, leading to suboptimal image quality
Solution Approach 1:
The patent merges data from multiple frames to reconstruct spatial images, exploiting both spatial and temporal redundancy. This approach improves image quality by utilizing all available information across multiple time points, while the use of efficient algorithms maintains computational feasibility despite the increased complexity of processing multiple frames together.
Data Source
AI summary
A spatio-temporal image of an object is reconstructed based on captured data characterizing the object. The spatio-temporal image comprises a plurality of spatial images in respective time intervals, and at least a given one of the spatial images in one of the time intervals is reconstructed using not only captured data from a frame associated with that time interval but also captured data associated with one or more additional frames associated with other time intervals. The spatio-temporal image may be reconstructed by iteratively obtaining a solution to a minimization or maximization problem in a sparse domain and transforming the solution to an image domain. The transformation between the sparse domain and the image domain may utilize a spatio-temporal transformation implemented using a plurality of basis functions, one or more of which may be determined at least in part based on secondary information associated with the imaged object.