Magnetic Resonance Image Reconstruction Using K-Space Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Generating image data of moving subjects, such as the heart, in magnetic resonance tomography is challenging due to movement artifacts caused by cardiac and breathing cycles, requiring longer measurement times and complex correction methods to achieve diagnostic-quality images.
Innovation Solution
A method and device that acquire raw data across different movement phases, reconstruct interim image data sets, determine deviation data to correct for movement, and reconstruct final image data using weighted raw data from various phases, allowing for reduced movement artifacts and shorter measurement times.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If raw data acquisition is triggered only in a defined movement phase (e.g., diastolic phase), then movement artifacts are reduced, but measurement time becomes excessively long
Solution Approach 1:
The patent segments k-space into multiple regions (central region and peripheral regions) and applies different acquisition strategies to each segment. The central region is acquired during the diastolic phase with triggering to minimize motion artifacts, while peripheral regions are acquired without strict triggering to reduce measurement time. This segmentation allows the system to balance image quality and measurement time by applying motion correction only where most critical.
2Area of stationary object
If raw data are acquired during different movement cycles to cover complete volume, then complete organ imaging is achieved, but significant movement artifacts occur
Solution Approach 1:
The patent divides the acquisition process into segments: central k-space lines are acquired during the diastolic phase with EKG triggering to ensure high quality and minimal motion artifacts, while peripheral k-space lines are acquired in subsequent cardiac cycles without strict triggering. This segmentation enables complete volume coverage while maintaining acceptable image quality through selective application of triggering.
Solution Approach 2:
The patent applies different quality standards to different regions of k-space. The central region, which contains the most important low-frequency information, is acquired with high quality constraints (triggering during diastole). Peripheral regions, containing high-frequency details, are acquired with relaxed constraints to reduce measurement time. This local quality approach optimizes the balance between complete coverage and image quality.
3Manufacturing precision
If triggering is applied to acquire data only in diastolic phase, then cardiac motion artifacts are reduced, but breathing movement causes heart position variation
Solution Approach 1:
The patent introduces a position correction mechanism as an intermediary step between data acquisition and final image reconstruction. Navigator echoes are acquired to monitor breathing-induced position changes, and these position corrections are applied to the raw data before reconstruction. This intermediary correction process compensates for heart position variations caused by breathing, maintaining reliability across multiple cardiac cycles.
Data Source
AI summary
In a method and apparatus for the generation of image data of a moving subject, raw data are initially acquired for a region comprising the subject at different measurement points in time in different movement phases of the subject. A reconstruction then takes place of multiple interim image data sets of the subject from the raw data that are respectively associated with different movement phases of the subject. Deviation data are then determined between the interim image data sets of the different movement phases of the subject, and the reconstruction of image data from raw data of different movement phases then takes place under consideration of the deviation data.