US-12625211-B2 - Model-based reconstruction for looping-star pulse sequences in MRI

US12625211B2US 12625211 B2US12625211 B2US 12625211B2US-12625211-B2

Abstract

The following relates generally to improved techniques for magnetic resonance imaging (MRI). In particular, the following relates to improving techniques of data acquisition and image reconstruction in relation to the “looping star” pulse sequence, which was developed to reduce acoustic noise. For example, techniques disclosed herein address the problem that looping star pulse sequence suffers from the overlapping echo effect. For instance, some implementations build echo-in and echo-out signal matrixes, and then sum the matrices to create a system matrix.

Inventors

Haowei Xiang
Jeffrey Allen Fessler
Douglas Clair Noll

Assignees

REGENTS OF THE UNIVERSITY OF MICHIGAN

Dates

Publication Date: 20260512
Application Date: 20230424

Claims (20)

1 . A method for use in T* 2 weighted magnetic resonance imaging (MRI) image reconstruction, the method comprising: collecting, via one or more processors, a plurality of echo signals, the plurality of echo signals being created by a MRI machine: (i) applying a gradient magnetic field to a subject, and (ii) applying a sequence of radio frequency (RF) pulses to the subject, wherein the plurality of echo signals include at least one overlap between echoes, and wherein the plurality of echo signals are collected in a phase during which the sequence of RF pulses are not being applied to the subject; building, via the one or more processors, an echo-in signal matrix based on k-space sampling locations; building, via the one or more processors, an echo-out signal matrix based on k-space sampling locations; summing, via the one or more processors, the echo-in signal matrix with the echo-out signal matrix to create a system matrix that models overlapping echoes; and reconstructing, via the one or more processors, a MRI image and/or rate map based on the system matrix, and the collected plurality of echo signals to resolve the overlapping echoes modeled in the system matrix.
2 . The method of claim 1 , further comprising: collecting, via the one or more processors, a free-induction-decay (FID) signal, the FID signal being a signal that is created by the MRI machine applying: (i) the gradient magnetic field to the subject, and (ii) the sequence of RF pulses to the subject, wherein the FID signal is collected in a phase during which the sequence of RF pulses are being applied to the subject; and building, via the one or more processors, a FID signal matrix based on the k-space sampling locations; and wherein the reconstruction of the MRI image and/or rate map is based further on the FID signal matrix, and or the collected FID signal.
3 . The method of claim 1 , wherein the reconstruction of the MRI image or rate map comprises reconstructing the rate map further based on a free-induction-decay (FID) signal.
4 . The method of claim 1 , further comprising: calculating, via the one or more processors, a series of optimized flip angles by maximizing a sum of transverse plane signals wherein the applying the sequence of RF pulses to the subject comprises applying the sequence of RF pulses such that the sequence of RF pulses induces the series of optimized flip angles; collecting, via the one or more processors, a free-induction-decay (FID) signal, the FID signal being a signal that is created by the MRI machine applying: (i) the gradient magnetic field to the subject, and (ii) the sequence of RF pulses to the subject, wherein the FID signal is collected in a phase during which the sequence of RF pulses are being applied to the subject; and building, via the one or more processors, a FID signal matrix based on the k-space sampling locations; and wherein the reconstruction of the MRI image and/or rate map is based further on the FID signal matrix, and/or the collected FID signal.
5 . The method of claim 1 , further comprising: calculating, via the one or more processors, a shaped sequence of RF pulses for more constant spatial excitation, wherein the applying the sequence of RF pulses comprises applying the shaped sequence of RF pulses; collecting, via the one or more processors, a free-induction-decay (FID) signal, the FID signal being a signal that is created by the MRI machine applying: (i) the gradient magnetic field to the subject, and (ii) the shaped sequence of RF pulses, wherein the FID signal is collected in a phase during which the shaped sequence of RF pulses are applied to the subject; and building, via the one or more processors, a FID signal matrix based on the k-space sampling locations; and wherein the reconstruction of the MRI image and/or rate map is based further on the FID signal matrix, a shaped excitation profile of the calculated shaped sequence of RF pulses, and/or the collected FID signal.
6 . The method of claim 1 , wherein: the building of the echo-in signal matrix further comprises applying weights to at least one row of the echo-in signal matrix; and the building of the echo-out signal matrix further comprises applying weights to at least one row of the echo-out signal matrix.
7 . The method of claim 1 , wherein the reconstruction of the MRI image and/or rate map is done according to an argument of the minimum equation that takes the system matrix, and the collected plurality of echo signals as inputs.
8 . The method of claim 1 , wherein the reconstruction of the MRI image and/or rate map is done according to an argument of the minimum equation with weighted elements.
9 . The method of claim 1 , wherein the echo-in signal matrix, and the echo-out signal matrix are built according to the equation: a lij =∫b ( {right arrow over (r)}−{right arrow over (r)} j ) c ( {right arrow over (r)} ) e −z({right arrow over (r)})t l,i e −l2π{right arrow over (k)} l (t l,i )·{right arrow over (r)} d{right arrow over (r)} where: l=1 builds the echo-in signal matrix; l=2 builds the echo-out signal matrix; b(⋅) is an object basis function; c({right arrow over (r)}) is a sensitivity map of a receiver coil; z({right arrow over (r)}) is a rate map; and e −i2π{right arrow over (k)} l (t l,i )·{right arrow over (r)} denotes a non-uniform fast Fourier transform (NUFFT).
10 . A device for use in T* 2 weighted magnetic resonance imaging (MRI) image reconstruction, the device comprising one or more processors configured to: collect a plurality of echo signals, the plurality echo signals being a plurality of signals that are created by a MRI machine: (i) applying a gradient magnetic field to a subject, and (ii) applying a sequence of radio frequency (RF) pulses to the subject, wherein the plurality of echo signals include at least one overlap between echoes, and wherein the one or more processors are further configured to collect in a phase during which the sequence of RF pulses are not being applied to the subject; build an echo-in signal matrix based on k-space sampling locations; build an echo-out signal matrix based on the k-space sampling locations; sum the echo-in signal matrix with the echo-out signal matrix to create a system matrix that models overlapping echoes; and reconstruct a MRI image and/or rate map based on the system matrix, and the collected plurality of echo signals to resolve the overlapping echoes modeled in the system matrix.
11 . The device of claim 10 , wherein the one or more processors are further configured to: collect a free-induction-decay (FID) signal, the FID signal being a signal that is created by the MRI machine applying: (i) the gradient magnetic field to the subject, and (ii) the sequence of RF pulses to the subject, wherein the one or more processors are configured to collect the FID signal in a phase during which the sequence of RF pulses are applied to the subject; and build a FID signal matrix based on the k-space sampling locations; and wherein the reconstruction of the MRI image and/or rate map is based further on the FID signal matrix, and/or the collected FID signal.
12 . The device of claim 10 , wherein the one or more processors are further configured to reconstruct the rate map further based on a free-induction-decay (FID) signal.
13 . The device of claim 10 , wherein the one or more processors are further configured to: calculate a series of optimized flip angles by maximizing a sum of transverse plane signals; apply the sequence of RF pulses to the subject by applying the sequence of RF pulses such that the sequence of RF pulses induces the series of optimized flip angles; collect a free-induction-decay (FID) signal, the FID signal being a signal that is created by the MRI machine applying: (i) the gradient magnetic field to the subject, and (ii) the sequence of RF pulses to the subject, wherein the one or more processors are configured to collect the FID signal in a phase during which the sequence of RF pulses are applied to the subject; and build a FID signal matrix based on the k-space sampling locations; and wherein the reconstruction of the MRI image and/or rate map is based further on the FID signal matrix, and/or the collected FID signal.
14 . The device of claim 10 , wherein the one or more processors are further configured to: calculate a shaped RF pulse for more constant spatial excitation; collect a free-induction-decay (FID) signal, the FID signal being a signal that is created by the MRI machine applying: (i) the gradient magnetic field to the subject, and (ii) the shaped RF pulse; and build a FID signal matrix based on the k-space sampling locations; and wherein the reconstruction of the MRI image and/or rate map is based further on the FID signal matrix, a shaped excitation profile of the calculated shaped sequence of RF pulses, and/or the collected FID signal.
15 . The device of claim 10 , wherein the one or more processors are further configured to: build the echo-in signal matrix further by applying weights to at least one row of the echo-in signal matrix; and build the echo-out signal matrix further by applying weights to at least one row of the echo-out signal matrix.
16 . The device of claim 10 , wherein the one or more processors are further configured to reconstruct the MRI image and/or rate map according to an argument of the minimum equation that takes the system matrix as an input.
17 . The device of claim 10 , wherein the one or more processors are further configured to reconstruct the MRI image and/or rate map according to an argument of the minimum equation with weighted elements.
18 . A magnetic resonance imaging (MRI) system for use in T* 2 weighted MRI image reconstruction, the MRI system comprising: one or more processors; a gradient coil; at least one radio frequency (RF) transmit coil; at least one RF receive coil; and one or more memories coupled to the one or more processors; the one or more memories including computer executable instructions stored therein that, when executed by the one or more processors, cause the one or more processors to: collect a plurality of echo signals, the plurality of echo signals being a plurality of signals that are created by the MRI system: (i) applying a gradient magnetic field to a subject, and (ii) applying a sequence of RF pulses to the subject, wherein the plurality of echo signals include at least one overlap between echoes, and wherein the one or more memories including computer executable instructions stored therein that, when executed by the one or more processors, further cause the one or more processors to collect in a phase during which the sequence of RF pulses are not being applied to the subject; build an echo-in signal matrix based on k-space sampling locations; build an echo-out signal matrix based on the k-space sampling locations; sum the echo-in signal matrix with the echo-out signal matrix to create a system matrix that models overlapping echoes; and reconstruct a MRI image and/or rate map based on the system matrix, and the collected plurality of echo signals to resolve the overlapping echoes modeled in the system matrix.
19 . The MRI system of claim 18 , wherein the one or more memories including computer executable instructions stored therein that, when executed by the one or more processors, further cause the one or more processors to: control the gradient coil to apply the gradient magnetic field to the subject; and control the RF transmit coil to apply the sequence of RF pulses including shaping the RF pulsed by placing a constraint on an amplitude of the RF pulses.
20 . The MRI system of claim 18 , wherein: the MRI image and/or rate map are included in a plurality of MRI images and/or a plurality of rate maps; and the plurality of MRI images and/or plurality of rate maps are reconstructed further by splitting time windows of the collected plurality of echo signals.

Description

CROSS REFERENCE TO RELATED APPLICATIONS This application claims the benefit of U.S. Provisional Application No. 63/333,831, entitled “Model-Based Reconstruction For Looping-Star Pulse Sequences In MRI” (filed Apr. 22, 2022), the entirety of which is incorporated by reference herein. STATEMENT OF GOVERNMENT INTEREST This invention was made with government support under EB023618 awarded by the National Institutes of Health. The government has certain rights in the invention. BACKGROUND In modern magnetic resonance imaging (MRI) systems, loud acoustic noises are generated from Lorentz forces caused by rapidly changing currents in the magnetic field gradient coils. To reduce these loud noises, a magnetic pulse sequence, known as the “looping star” pulse sequence, was developed. However, the looping star pulse sequence has several drawbacks. For example, the looping star pulse sequence suffers from an overlapping echo effect, which comes from signals from multiple excitation pulses being simultaneously present while looping through k-space locations. The systems and methods disclosed herein provide solutions to this problem and others. SUMMARY This Summary is provided to introduce a selection of concepts in a simplified form that are further described below in the Detailed Description. This Summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter. In one aspect, a method for use in T2* -weighted magnetic resonance imaging (MRI) image reconstruction may be provided. The method may include: (1) collecting, via one or more processors, a plurality of echo signals, the plurality of echo signals being created by a MRI machine: (i) applying a gradient magnetic field to a subject, and (ii) applying a sequence of radio frequency (RF) pulses to the subject, wherein the plurality of echo signals are collected in a phase during which the sequence of RF pulses are not being applied to the subject; (2) building, via the one or more processors, an echo-in signal matrix based on k-space sampling locations; (3) building, via the one or more processors, an echo-out signal matrix based on the k-space sampling locations; (4) summing, via the one or more processors, the echo-in signal matrix with the echo-out signal matrix to create a system matrix that models overlapping echoes; and (5) reconstructing, via the one or more processors, a MRI image and/or rate map based on the system matrix, and the collected plurality of echo signals. In another aspect, a device for use in T2* weighted magnetic resonance imaging (MRI) image reconstruction may be provided. The device may include one or more processors configured to: (1) collect a plurality echo signals, the plurality of echo signals being a plurality of signals that are created by a MRI machine: (i) applying a gradient magnetic field to a subject, and (ii) applying a sequence of radio frequency (RF) pulses to the subject, wherein the one or more processors are further configured to collect in a phase during which the sequence of RF pulses are not being applied to the subject; (2) build an echo-in signal matrix based on k-space sampling locations; (3) build an echo-out signal matrix based on the k-space sampling locations; (4) sum the echo-in signal matrix with the echo-out signal matrix to create a system matrix that models overlapping echoes; and (5) reconstruct a MRI image and/or rate map based on the system matrix, and the collected plurality of echo signals. In yet another aspect, a system for use in T2* -weighted magnetic resonance imaging (MRI) image reconstruction may be provided. The system may include: (a) one or more processors; (b) a gradient coil; (c) at least one radio frequency (RF) transmit coil; (d) at least one RF receive coil; and (e) one or more memories coupled to the one or more processors. The one or more memories may include computer executable instructions stored therein that, when executed by the one or more processors, may cause the one or more processors to: (1) collect a plurality of echo signals, the plurality of echo signals being a plurality of signals that are created by a MRI machine: (i) applying a gradient magnetic field to a subject, and (ii) applying a sequence of RF pulses to the subject, wherein the one or more memories including computer executable instructions stored therein that, when executed by the one or more processors, further cause the one or more processors to collect in a phase during which the sequence of RF pulses are not being applied to the subject; (2) build an echo-in signal matrix based on k-space sampling locations; (3) build an echo-out signal matrix based on the k-space sampling locations; (4) sum the echo-in signal matrix with the echo-out signal matrix to create a system matrix that models overlapping echoes; and (5) reconstruct a MRI image and/or rate map based on the system matrix, and the collected