MRI Data Processing Using Preliminary Unfolding and Iterative Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional Magnetic Resonance Imaging (MRI) methods using parallel imaging (PI) and Half Fourier Imaging (HFI) together require repeated data processing on a channel-by-channel basis, leading to increased data processing time and complexity.
Innovation Solution
A magnetic resonance imaging apparatus that includes a data collection unit for collecting data using a Half Fourier method with multiple coils, an unfolding unit for generating unfolded image data, and a data processing unit that performs repeated data filling and phase correction processes to improve accuracy, allowing for reduced data processing time by performing these processes after initial unfolding.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If repeated data filling and phase correction processes are performed on a channel-by-channel basis using conventional PI and HFI methods, then image reconstruction accuracy is improved, but data processing time increases significantly
Solution Approach 1:
The patent performs the unfolding process preliminarily before the repeated data filling and phase correction processes. By completing the unfolding step first, the subsequent iterative HFI processes operate on already-unfolded data, reducing the computational burden and processing time required for each iteration while maintaining accuracy improvement.
Solution Approach 2:
The patent segments the data processing into distinct stages: first performing the unfolding process separately, then performing the data filling and phase correction processes. This segmentation allows the complex iterative HFI processes to operate more efficiently on pre-processed data, reducing overall processing time while maintaining accuracy.
2Productivity
If multiple coil elements are used for parallel imaging, then imaging speed is improved, but data processing complexity increases
Solution Approach 1:
The patent performs the unfolding process as a preliminary step before the iterative HFI processes. This preliminary unfolding step simplifies the subsequent data filling and phase correction operations by working with already-unfolded intermediate image data, thereby reducing the computational complexity associated with processing data from multiple coil elements.
Solution Approach 2:
The patent introduces unfolded intermediate image data as an intermediary representation between the raw k-space data from multiple coils and the final image data. This intermediary format simplifies the subsequent processing steps by providing a more manageable data structure that reduces computational complexity while preserving the benefits of parallel imaging.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach significantly reduces data processing time and complexity while maintaining high accuracy in image reconstruction, allowing for faster and more efficient MRI data acquisition.
Implementation Method 1
Magnetic resonance imaging is an imaging method which magnetically excites nuclear spins of a patient placed in a static magnetic field with an RF signal at the Larmor frequency to reconstruct an image using a magnetic resonance signal resulting from the excitation
Data Source
AI summary
According to one embodiment, a magnetic resonance imaging apparatus includes; a data collection unit which collects magnetic resonance data from a patient by a half Fourier method using a plurality of coils; an unfolding unit which performs an unfolding process on a plurality of items of folded image data obtained from the plurality of coils to generate unfolded image data by using sensitivity data of the plurality of coils; and a data processing unit which repeatedly performs a data filling process and a phase correction process to improve accuracy of data in an unsampled region to generate image data for display, the data filling process filling the unsampled region in k-space with k-space data obtained by Fourier-transforming the unfolded image data and the unsampled region being a region for which the data have not been collected.