Time-Domain MRI Signal Reconstruction via Iterative Parameter Fitting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current time-domain magnetic resonance imaging techniques lack efficiency in reconstructing spatial parameter distributions of magnetization and material properties, leading to suboptimal image quality and information extraction.
Innovation Solution
A system and method that includes a radio frequency excitation device, a receiving coil, and a processor to generate a simulated signal in the time domain based on spatial parameter distributions, such as magnetization and electromagnetic field distributions, and fit these distributions to accurately match the received MRI signal, thereby improving image reconstruction and information extraction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional inverse FFT and k-space data with Cartesian sampling are used for MRI image reconstruction, then the reconstruction process is efficient and straightforward, but the ability to extract quantitative parameter maps (such as T1, T2, R2*) is limited and requires separate multi-echo acquisitions
Solution Approach 1:
The patent transforms the reconstruction approach from spatial domain (FFT-based) to temporal domain fitting, changing the fundamental parameter space. By fitting the temporal signal evolution directly to Bloch equations, the method simultaneously extracts multiple quantitative parameters (T1, T2, R2*, field map) from a single acquisition, converting a limitation into an advantage
Solution Approach 2:
The temporal domain fitting algorithm serves multiple functions simultaneously: it reconstructs the image, quantifies T1 relaxation, quantifies T2 relaxation, measures R2* decay, and corrects off-resonance effects. This multi-functional approach eliminates the need for separate acquisitions for each parameter, directly addressing the information loss problem
2Productivity
If single-shot MRI sequences are used to reduce acquisition time, then scanning speed improves, but the accuracy of parameter estimation deteriorates due to insufficient signal evolution time
Solution Approach 1:
The patent employs iterative fitting algorithms that use the acquired signal to generate initial parameter estimates, then refine these estimates by comparing predicted signals (from Bloch equations) with actual measured signals. This feedback loop continues until convergence, allowing accurate parameter extraction even from single-shot data where signal evolution is limited
Solution Approach 2:
The method performs preliminary signal modeling using Bloch equations before final parameter extraction. By pre-calculating expected signal evolution for different parameter combinations and comparing against actual data, the system prepares the optimal parameter estimates in advance, improving accuracy without requiring additional scan time
3Measurement precision
If iterative algorithms are used for off-resonance correction and parameter estimation, then measurement accuracy improves, but processing time and computational complexity increase significantly
Solution Approach 1:
The patent merges off-resonance correction, T1 estimation, T2 estimation, and R2* measurement into a single unified temporal domain fitting procedure. By combining these previously separate iterative processes into one simultaneous optimization problem, the method achieves comparable or superior accuracy while reducing total processing time through algorithmic efficiency
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 enhances the accuracy of spatial parameter distribution fitting, providing valuable information about material properties like relaxation times and electromagnetic fields, leading to improved image reconstruction and medical application insights.
Implementation Method 1
transiently exciting a sample thereby causing the sample to emit an MRI signal
Implementation Method 2
a receiving coil for receiving the MRI signal
Implementation Method 3
a receiving coil for receiving the MRI signal; wherein the received MRI signal is a time domain signal free from Fourier transform
Data Source
Figure 1~2
Figure 3
Figure 4~5
AI summary
A system for performing time-domain magnetic resonance imaging comprises an excitation device for transiently exciting a sample thereby causing the sample to emit an MRI signal. A receiving coil receives the MRI signal. A simulated signal of the receiving coil is generated in a time domain, based on a plurality of spatial parameter distributions, wherein the spatial parameter distributions include a spatial distribution of a magnetization, wherein the spatial parameter distributions further include at least one of a spatial distribution of a material property of a material of the sample and a spatial distribution of an electromagnetic field. An objective function is based on a difference between the received MRI signal and the simulated signal in the time domain. The plurality of spatial parameter distributions are fitted based on the objective function. The sample is excited again before the sample reaches an equilibrium state.