Pulse Sequence Database for MRI B1 Homogeneity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-field magnetic resonance imaging (MRI) systems face challenges with B1 artifacts, particularly at 3T and above, leading to inhomogeneous nuclear magnetization and reduced diagnostic accuracy due to existing RF shimming methods' limitations in achieving homogeneous excitation across various subjects, especially in the abdomen.
Innovation Solution
A method involving clustering of subjects based on morphological features using machine learning to design pseudo-universal pulse sequences, which are optimized for each cluster, allowing for robust and efficient assignment of the best pulse sequence to individual subjects without requiring extensive calibration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If static active RF shimming is used to homogenize the B1 field, then the B1 field homogeneity is improved for most patients, but it fails to provide sufficient homogeneity in 10 to 20% of cases and is not satisfactory for higher field values
Solution Approach 1:
The patent transitions from static RF shimming to dynamic RF shimming using kT-points method, where multiple RF coil elements transmit RF pulses with different time-varying complex envelopes. This dynamic approach allows continuous adjustment of RF parameters during the pulse sequence to maintain homogeneous excitation across all patient types and field strengths.
Solution Approach 2:
The patent changes the RF pulse parameters by using different time-varying complex envelopes for each RF coil element in the parallel transmission system. By optimizing these temporal parameters dynamically, the system achieves homogeneous B1 field distribution that static weighting cannot provide.
2Manufacturing precision
If dynamic RF shimming with kT-points is used to achieve better uniformity, then the excitation homogeneity is improved, but the complexity increases making it essentially a research tool
Solution Approach 1:
The patent performs preliminary calibration to measure B1+ maps and off-resonance frequency maps from each transmit channel before the actual imaging sequence. These pre-measured parameters are stored and reused, eliminating the need for repeated complex calculations during clinical operation and reducing system complexity.
Solution Approach 2:
The patent creates a database of pre-computed kT-points pulse sequences with optimized complex envelopes for different imaging scenarios. These pre-computed pulse designs are copied and applied during clinical practice, avoiding repeated complex optimization calculations and simplifying the system implementation.
3Manufacturing precision
If calibration is performed to optimize weighting coefficients or RF waveforms, then the B1 field homogeneity is improved, but the calibration process is time consuming lasting nearly two minutes
Solution Approach 1:
The patent performs the time-consuming calibration process (B1+ mapping, off-resonance mapping, and pulse design optimization) before the actual imaging sequence. These pre-computed parameters are stored in a database and rapidly retrieved during clinical operation, separating the time-consuming optimization from the time-critical imaging process.
Solution Approach 2:
The patent implements a dynamic workflow where the system adapts between offline calibration mode (for optimization) and online imaging mode (for rapid execution). This dynamic separation allows comprehensive calibration without extending clinical scan time, as the optimized parameters are pre-determined and quickly applied.
4Ease of operation
If universal pulses are used for calibration-free dynamic RF-shimming, then the ease of operation is improved, but they fail to provide homogeneous enough excitation in a significant number of cases
Solution Approach 1:
The patent segments the patient population into different clusters based on morphological features (body size, shape, anatomy). Each cluster receives a tailored pulse sequence optimized for its specific characteristics, rather than applying a single universal pulse to all patients. This segmentation maintains ease of operation while improving excitation homogeneity for each subgroup.
Solution Approach 2:
The patent applies the principle of local quality by providing different pulse sequence characteristics tailored to specific patient clusters. Each cluster receives locally optimized pulses matched to its morphological features, rather than a homogeneous universal approach, thereby achieving both operational simplicity and excitation precision.
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 improves the homogeneity of magnetization flip angles, reducing the impact of B1 artifacts and enhancing diagnostic accuracy by providing robust MRI sequences less sensitive to subject movement and other perturbations, while eliminating the need for time-consuming calibration processes.
Implementation Method 1
manipulating the nuclear magnetization of a sample immersed in a static magnetic field, resulting from the orientation of nuclear spins
Implementation Method 2
A pulse sequence comprises one or more radio-frequency (RF) pulses and at least one magnetic field gradient waveform, allowing manipulating the nuclear magnetization
Data Source
Figure 1~2
Figure 3a~5
Figure 4a~4e
AI summary
A computer-implemented method of building a database of pulse sequences for parallel-transmission magnetic resonance imaging, comprising: a) for each of a plurality (S0) of subjects, determining an optimal sequence for the subject; b) for each subject, computing the values of said or of a different cost or merit function obtained by playing the optimal sequences for all the subjects; c) aggregating the subjects into a plurality of clusters using a clustering algorithm taking said values, or functions thereof, as metrics; d) for each cluster, determining an averaged optimal sequence for the cluster; e) receiving, as input, a set of features characterizing an imaging subject (IS), comprising at least a morphological feature of the subject; f) associating the subject to one pulse sequence of the database based on said set of features using the computer-implemented classifier algorithm; and g) performing magnetic resonance imaging using said pulse sequence. A magnetic resonance imaging apparatus for carrying out steps e) - g) of such a method.