Method and system for simultaneous mapping of quantitative MRI parameters using a T2 prepared inversion
Inventors
Bonanno, Gabriele • Marques, Jose Pedro • Kober, Tobias • HILBERT, Tom
Assignees
Siemens Healthineers AG • Stichting Radboud Universitair Medisch Centrum
Publication Number
US-12336802-B2
Publication Date
2025-06-24
Expiration Date
2043-04-14
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Abstract
A qMRI system and method map qMRI parameters of a biological object. The method includes performing, by the qMRI system, N scans wherein each scan, includes: performing T2-prepared inversion pulse series, each followed by readout blocks, each magnetization preparation RF pulse series is a T2-prepared inversion pulse series containing multiple pulses and an inter-pulse duration, for varying to obtain different T2 weightings; and acquiring, by the MRI system and during each readout block of an MRI, a recovery signal generated by a part of the biological object, wherein for each readout block, an MRI signal is acquired by the MRI system at different inversion times. An image of the part is reconstructed for and from each MRI signal. A voxel-wise signal is created by concatenating intensity values for a same voxel for the reconstructed images. A physical model is fitted to the concatenated intensity values to obtain a qMRI map.
Core Innovation
The invention relates to a method and system for simultaneous mapping of quantitative MRI (qMRI) parameters such as T1 and T2 relaxation times of a biological object using a T2-prepared inversion magnetization preparation. The method involves performing multiple MRI scans each including T2-prepared inversion pulse series followed by multiple readout blocks at varying inversion times, generating MRI signals with combined T1 and T2 weighting components. The MRI signals are reconstructed into images for each readout, and voxel-wise signals are created by concatenating intensity values for the same voxel across images acquired at different inversion times and echo times. A physical model is then fitted to these concatenated values to generate quantitative MRI parameter maps.
The problem being solved arises from limitations of existing qMRI methods which either require long acquisition times, suffer from experimental bias, or are designed to estimate only a single parameter at a time. Conventional methods for T1 and T2 mapping such as inversion-recovery sequences or T2-prepared rapid gradient echoes have lengthy acquisition times and are not readily suitable for clinical settings with high spatial resolution and large volume coverage. Alternative methods like DESPOT1/2 and MR fingerprinting enable simultaneous parameter estimation but suffer from potential bias and rely on complex sequences less supported by existing MRI hardware. The invention aims to achieve accurate, efficient simultaneous mapping of multiple qMRI parameters overcoming these limitations.
The invention achieves this by employing a novel combination of a T2-prepared inversion pulse sequence, specifically an adiabatic pulse series containing tip-down and refocusing pulses, which encodes T2 weighting into the longitudinal magnetization inversion and subsequent T1 recovery. Unlike prior methods, multiple readout blocks (at least two) at different inversion times are acquired following each magnetization preparation, allowing rich encoding of both T1 and T2 contrast. The MRI system varies the echo time (inter-pulse duration) within or between scans to obtain different T2 weighting levels. The reconstructed images are used to create voxel-wise signals that incorporate these dual-contrast information, which are quantitatively fitted to physics-based models or dictionaries (using Bloch equations or extended-phase-graph simulations) to produce quantitative maps of tissue parameters in a manner compatible with clinical scanner hardware and fast acquisition.
Claims Coverage
The independent claims define inventive features of a quantitative MRI method and system for simultaneous mapping of qMRI parameters using a T2-prepared inversion pulse series with multiple readout blocks per magnetization preparation.
Simultaneous mapping using T2-prepared inversion pulse series
Performing N scans each comprising one or more magnetization preparation RF pulse series, each followed by M_j readout blocks. Each magnetization preparation RF pulse series is a T2-prepared inversion containing multiple pulses characterized by an echo time (inter-pulse duration) TEp_j that is varied to obtain different T2 weightings. The pulse series inverts longitudinal magnetization while encoding T2 contrast during subsequent T1 recovery.
Acquisition of MRI recovery signals at multiple inversion times
During each readout block, MRI signals S_i,j are acquired at different inversion times TI_i,j with M_j≥2, allowing multiple images per magnetization preparation at different TIs.
Image reconstruction and voxel-wise signal creation
Reconstruction of an image I_i,j from each MRI signal, followed by creating a voxel-wise signal by concatenating intensity values corresponding to the same voxel across all reconstructed images acquired at varying (TI_i,j; TEp_j) pairs.
Fitting a physical model to voxel-wise signals for qMRI mapping
Fitting a physics-based model to the concatenated voxel-wise intensity values to obtain quantitative maps of one or more qMRI parameters.
Dictionary fitting procedure with numerical simulation
Performing numerical simulations generating sets of simulated voxel-wise signals associated to qMRI parameters to create a dictionary, and reconstructing quantitative maps by matching acquired voxel-wise signals to simulated ones to assign parameter values.
Use of an adiabatic T2-prepared inversion pulse series
The T2-prepared inversion pulse series may be adiabatic, comprising two rectangular tip-down RF pulses temporally framing two hyperbolic-secant-refocusing adiabatic RF pulses, with the inter-pulse duration measured between the centers of the first and fourth pulses.
Incorporation of additional parameter maps for fitting
Using additional acquired maps, such as B1+, B1-, B0 frequency offset, or diffusion maps, to select the subset of simulated signals for matching to reduce bias.
Physical model choices and qMRI parameters
The physical model used for fitting may include extended-phase-graph, Bloch-Equations, or analytical solutions to estimate parameters like T1, T2, T2*, B1+, proton density, magnetization transfer, diffusion, perfusion, fat/water fraction, and others.
The independent claims collectively cover a novel quantitative MRI method and system that utilize a T2-prepared inversion magnetization preparation with multiple readouts per preparation, acquisition of MRI signals at varied inversion and echo times, creation of voxel-wise concatenated signals, and fitting of physics-based models or dictionaries to reconstruct simultaneous qMRI parameter maps with improved accuracy, efficiency, and compatibility with clinical hardware.
Stated Advantages
Simultaneous estimation of multiple qMRI parameters (e.g., T1 and T2) with reduced measurement bias by including both relaxation mechanisms in the signal model.
Shorter acquisition time and minimized experimental bias compared to conventional methods requiring separate parameter estimation.
Use of a simple signal model and conventional acquisition technology (Cartesian k-space sampling, magnetization preparation, FLASH readout) facilitates compatibility with existing scanner hardware for clinical use.
Capability for high isotropic spatial resolution imaging with large volumetric coverage.
Lower sensitivity to B1+ transmit field inhomogeneity due to low flip-angle readouts.
Lower specific absorption rate (SAR) compared to other T2 mapping methods like multi-echo spin echo sequences.
Documented Applications
Mapping tissue-specific and organ-specific quantitative MRI parameters for use in diagnosis, therapy monitoring, and clinical interventions.
High isotropic resolution T1 and T2 mapping of biological objects such as the brain.
Clinical quantitative MRI examinations requiring simultaneous multi-parameter mapping with improved accuracy and efficiency.
Interested in licensing this patent?