Medical Image Reconstruction via Parallel Subset Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Medical imaging technologies face delays in workflow due to the time-consuming process of generating high-resolution images using compressed sensing, which can interrupt the scanning process and hinder immediate confirmation of data capture by operators.
Innovation Solution
A method involving the generation of a lower-resolution image using a subset of scan data, followed by a higher-resolution image using the full data set, allowing for faster initial image confirmation and reduced workflow interruptions while still providing high-resolution images for diagnosis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If compressed sensing reconstruction methods are used to generate high-resolution images, then image quality is improved, but reconstruction time increases significantly
Solution Approach 1:
The patent segments the scan data into multiple subsets and performs parallel reconstruction processes on each subset using different reconstruction methods. This divides the computationally intensive full reconstruction into smaller, manageable tasks that can be executed concurrently, reducing overall reconstruction time while maintaining image quality through selective combination of results.
Solution Approach 2:
The patent performs multiple reconstruction processes using different portions of the scan data subsets. By using partial data with different reconstruction algorithms and combining results, the system achieves high-quality reconstruction without processing the entire dataset through the most computationally expensive methods, thus reducing reconstruction time.
2Measurement precision
If full scan data is processed for high-resolution image generation, then diagnostic accuracy is improved, but workflow efficiency deteriorates due to interruptions
Solution Approach 1:
The patent performs preliminary reconstruction processes on subsets of scan data in parallel while the scanning is still in progress or immediately upon completion. This preliminary action provides initial image results that allow operators to confirm data capture and plan subsequent scans without waiting for the complete high-resolution reconstruction, thereby maintaining workflow efficiency.
Solution Approach 2:
The patent enables continuous workflow by performing reconstruction processes that do not interrupt the scanning operation. Multiple reconstruction processes run in parallel or concurrently with data acquisition, ensuring that the useful action of image generation continues without stopping the scanning process, thus maintaining productivity.
3Manufacturing precision
If iterative reconstruction methods are used, then image resolution is improved, but computational cost increases
Solution Approach 1:
The patent segments the computational task into multiple parallel reconstruction processes working on different data subsets. By distributing the iterative reconstruction computations across multiple processes that can run concurrently, the total computational cost is reduced while maintaining high image resolution through the use of iterative methods on smaller, manageable data portions.
Solution Approach 2:
The patent changes computational parameters by using different reconstruction methods with varying computational intensities on different data subsets. Some subsets undergo full iterative reconstruction for maximum resolution, while others use faster methods, optimizing the balance between image resolution and computational cost through parameter variation across the dataset.
Data Source
AI summary
Devices, systems, and methods for generating a medical image receive scan data, wherein the scan data is defined in an acquisition space; perform a first reconstruction process to generate a lower-resolution image based on the scan data, wherein the first reconstruction process uses a subset of the scan data; and perform a second reconstruction process to generate a higher-resolution image based on the scan data, wherein the second reconstruction process uses more of the scan data than the subset of the scan data.