Magnetic resonance imaging using frequency-modulated radio frequency pulses for frequency encoding of magnetic resonance signals

FE-FREE addresses the accessibility and cost issues of MRI systems by using B1 field gradients for spatial encoding, enhancing imaging speed and reducing infrastructure needs, making MRI more accessible globally.

WO2024243156A1PCT designated stage expired Publication Date: 2024-11-28REGENTS OF THE UNIVERSITY OF MINNESOTA

Patent Information

Application Number
PCT/US2024/030266
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-05-19
Filing Date
2024-05-20
Publication Date
2024-11-28

AI Technical Summary

Technical Problem

Current MRI systems are costly and inaccessible to 90% of the world's population due to their high cost and complexity, with B1 encoding techniques being limited by slow acquisition speed and sensitivity to B0 off-resonance.

Method used

The implementation of frequency-encoding frequency-modulated Rabi encoded echo (FE-FREE) methods, which utilize B1 field gradients produced by frequency-modulated radiofrequency pulses to spatially encode magnetic resonance signals without the need for B0 gradient coils, enabling imaging in inhomogeneous B0 and B1 fields.

Benefits of technology

FE-FREE allows for faster and more efficient spatial encoding of MRI signals, reducing the need for costly infrastructure and enabling imaging in challenging field conditions, thus making MRI more accessible and affordable.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure IMGF000005_0001
    Figure IMGF000005_0001
  • Figure IMGF000006_0001
    Figure IMGF000006_0001
  • Figure IMGF000006_0002
    Figure IMGF000006_0002
Patent Text Reader

Abstract

A frequency-modulated magnetic resonance imaging (MRI) technique uses frequency-modulated radio frequency (RF) pulses to provide spatial encoding without magnetic field gradients. Frequency encoding is provided by using a first frequency- modulated RF pulse to dephase spins. A second frequency-modulated RF pulse is then used to rephase, or refocus, the dephased spins. While the second frequency-modulated RF pulse is generated, k -space data are acquired by sampling one or more echo signals formed during the application of the second frequency-modulated RF pulse. Images may be reconstructed from the k-space data. In some examples, k-space data are transformed into a rotating reference frame using a rotation matrix and images may be reconstructed from the transformed k-space data.
Need to check novelty before this filing date? Find Prior Art

Description

UMN 2022-135Quarles Ref.920171.00582MAGNETIC^RESONANCE^IMAGING^USING^FREQUENCY‐MODULATED^RADIO^ FREQUENCY^PULSES^FOR^FREQUENCY^ENCODING^OF^MAGNETIC^RESONANCE^ SIGNALS^ CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Patent ApplicationSerial No. 63 / 503,420, filed on May 19, 2023, and entitled “MAGNETIC RESONANCEIMAGING USING FREQUENCY-MODULATED RADIO FREQUENCY PULSES FORFREQUENCY ENCODING OF MAGNETIC RESONANCE SIGNALS,” which is hereinincorporated by reference in its entirety.STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH

[0002] This invention was made with government support under EB026061 andEB025153 awarded by the National Institutes of Health. The government has certainrights in the invention.BACKGROUND

[0003] The high cost of magnetic resonance imaging (MRI) systems relegates theirclinical impact to the world’s wealthiest communities. Because standard MRI technologyis inaccessible to approximately 90% of the world’s population, there is a global unmetneed to develop a greener, cheaper, and smaller MRI systems.

[0004] B1 encoding techniques aim to remove the need for B0 gradient coils, whichreduce the cost, complexity, and infrastructure demands of an MRI system. B1 imagingwas first developed with Rotating Frame Zeugmatography, but it was never adopted dueto its sensitivity to B0 off-resonance and slow acquisition. A resurgence of B1 encodingtechniques has emerged. In particular, phase-encoding frequency-modulated Rabiencoded echo (PE-FREE) is a phase-encoding B1 approach capable of imaging in highlyinhomogeneous B0 and B1 fields. Despite the potential of these B1 imaging approaches,these phase-encoding techniques are limited by the slow speed of the acquisition andencoding.SUMMARY OF THE DISCLOSURE

[0005] The present disclosure addresses the aforementioned drawbacks by 1 QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582providing a system and method for frequency-encoding frequency-modulated Rabiencoded echo (FE-FREE), which utilizes B1 field gradients produced by frequency-modulated radiofrequency pulses paired with nearly simultaneous RF transmit andreceive to achieve frequency encoding to spatially encode the signal. The disclosedmethods can generate rotary echoes without the use of B0 gradient coils and can produceimages in the midst of B0 inhomogeneity.

[0006] It is an aspect of the present disclosure to provide a method for spatiallyencoding magnetic resonance signals during magnetic resonance imaging (MRI). Themethod includes generating, with an MRI system, a radio frequency (RF) excitation pulseto excite spins in a region-of-interest. A first frequency-modulated RF pulse is generatedwith the MRI system to dephase the spins excited by the RF excitation pulse. A secondfrequency-modulated RF pulse is then generated with the MRI system to rephase thespins dephased by the first frequency-modulated RF pulse. Additionally, k-space data areacquired by sampling at least one echo signal generated while rephasing the spins withthe second frequency-modulated RF pulse.

[0007] It is another aspect of the present disclosure to provide a system forspatially encoding magnetic resonance signals during magnetic resonance imaging (MRI).The system includes a magnet system configured to produce a main magnetic field and aradio frequency (RF) coil arrangement. The RF coil arrangement generates an RFexcitation pulse to excite spins in a region-of -interest; generates a first frequency-modulated RF pulse to dephase the spins excited by the RF excitation pulse; generates asecond frequency-modulated RF pulse to rephase the spins excited by the RF excitationpulse; and receives at least one echo signal generated while rephasing the spins with thesecond frequency-modulated RF pulse. The first frequency-modulated RF pulse isspatially varying in the region-of-interest along a first dimension, and the secondfrequency-modulated RF pulse is spatially varying in the region-of-interest along the firstdimension.

[0008] These are but a few, non-limiting examples of aspects of the presentdisclosures. Other features, aspects and implementation details will be describedhereinafter.BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Various objects, features, and advantages of the disclosed subject matter 2 QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582can be more fully appreciated with reference to the following detailed description of thedisclosed subject matter when considered in connection with the following drawings, inwhich like reference numerals identify like elements.

[0010] FIG. 1 demonstrates an example of phase accrual achieved using a B1gradient in accordance with some aspects of the present disclosure.

[0011] FIG. 2A shows an example of a pulse sequence that can be used to achievespatial encoding by generating a B1 field in accordance with some aspects of the presentdisclosure.

[0012] FIG.2B shows an example of a pulse sequence that includes sampling of anecho produced by a B1 field in accordance with some aspects of the present disclosure.

[0013] FIG.2C shows another example of a pulse sequence that includes samplingof an echo produced by a B1 field in accordance with some aspects of the presentdisclosure.

[0014] FIG. 3A shows an example of k-space sampling that can be achieved withthe methods disclosed herein.

[0015] FIG. 3B shows another example of k-space sampling that can be achievedwith the methods disclosed herein.

[0016] FIG.4A shows an example of the effective nutation frequency achieved witha B1 field in accordance with some aspects of the present disclosure.

[0017] FIG. 4B shows another example of the effective nutation frequencyachieved with a B1 field in accordance with some aspects of the present disclosure.

[0018] FIG. 5 shows an example of a trajectory of spins produced by the methodsdisclosed herein.

[0019] FIG. 6 shows a plot of magnetization components in a rotating frameproduced by the methods disclosed herein.

[0020] FIG.7 shows an example of magnetization components in a first and secondrotating frame in accordance with some aspects of the present disclosure.

[0021] FIG. 8 shows a flow chart that illustrates a process to perform methodsdescribed in the present disclosure.

[0022] FIG.9 shows an example of a pulse sequence that can be used to achieve 2Dencoding by combining PE-FREE and FE-FREE in accordance with some aspects of thepresent disclosure.

[0023] FIG. 10 shows example data demonstrating simulated rotary echoes 3 QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582produced in accordance with some aspects of the present disclosure.

[0024] FIG. 11 shows an example of adiabatic full passage pulses simulated forvarying bandwidth, which may be used in accordance with the present disclosure.

[0025] FIG. 12 shows experimental results demonstrating the signal acquiredusing methods described in the present disclosure.

[0026] FIG.13 shows an example experimental setup.

[0027] FIG. 14 shows simulation results demonstrating the signal acquired usingmethods described in the present disclosure.

[0028] FIG.15 shows an example of experimental results obtained using PE-FREEin accordance with the present disclosure.

[0029] FIG.16 shows an example of experimental results obtained using FE-FREEin accordance with the present disclosure.

[0030] FIG. 17 is a block diagram of an example magnetic resonance imaging(“MRI”) system that can implement the methods described in the present disclosure.

[0031] FIG. 18 is a block diagram of an example MRI system that can implementthe methods of the present disclosure.

[0032] FIG. 19 is a block diagram of example components that can implement thesystem of FIG.18.DETAILED DESCRIPTION

[0033] B1 imaging approaches (e.g., frequency-modulated Rabi encoded echo(FREE)) aim to develop cheaper, quieter, and smaller MRI systems. FREE techniques canachieve spatial encoding by producing a radio-frequency (RF) induced phase pattern,circumventing the need for B0 gradient systems. The FREE technique can include the useof adiabatic full passage (AFP) pulses to produce RF-induced phase patterns in space.

[0034] AFPs are able to uniformly invert spins across inhomogeneous B0 and B1fields. AFPs can be defined by their amplitude and frequency modulation functionsdescribed by AM(t) and FM(t), respectively. AM(t) can modulate the pulse amplitude (ω1)according to:AM ^ t ^ ^ ^ 1( t ), where^ 1 ^ ^B 1. FM(t) can modulate the transmit frequency(ωRF) of the(ωc), which is typically set to the Larmorfrequency, according to: FM ^ t ^ ^ ^c ^ ^ RF ^ t ^ . An AFP can further be described by its time-bandwidth product, defined as R = Tp∙BW ^, where Tp is the pulse duration and BW ^ is the4 QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582bandwidth of the frequency-modulated pulse.

[0035] During the application of an AFP, AM(t) and FM(t) may define fieldsproduced in the x ^ and z ^ directions, for example, described by: ^1(t ) ^ ^ B 1 ^ t ^ x ^ ' and^ ^(t ) ^ ^ ^c ^ ^ RF ^ t ^ ^ z ^' . The vector sum of these produced fields defines the nutationrotating direction during the application of an AFP pulse defined as:ω^t^ ൌ ω ^ ^ ᇱ ^ ^ ᇱ^^^ ^ t ^^ ^ Δω t ^^ (1).

[0036] The effective frequency, ω^^^^t^, sweeps from the +z axis to the –z axis,tipping at the angular velocity, which can be defined as ^^ ^ rctan ^ t ^^ t ^a ^ 1 ^ ^.^ ^ ^ t ^^ ^ ^ ^ ^

[0037] The adiabatic condition states that if |ω^^^^t^| ≫ then spins thatbegin collinear or perpendicular with ω^^^^t^ remain so while ω^^^^t^ sweeps. Theadiabatic condition enables uniform inversion in highly inhomogeneous B0 and B1 fields.Such pulses that satisfy the adiabatic condition can be used for frequency-encoded FREE(FE-FREE), as disclosed herein.

[0038] As a non-limiting example, hyperbolic secant (HS) pulses are one class offrequency-modulated pulses, which provide smooth inversion profiles within thebandwidth of the pulse (BWΩ). The AM(t) and FM(t) of an HS pulse may be written as:^1^t^ ^^max ^1sech ^ ^^^2 t ^n^^ 1 ^ ^(2)(3)

[0039] isthe pulse order, ^^ is a dummy integration variable, and A = BWΩ / 2. HS pulses aredescribed herein as a non-limiting example for demonstration purposes. However, otherfrequency-modulated pulses can be employed, such as BIR4, chirp, etc.

[0040] In the disclosed methods and systems, AFPs or other frequency modulatedpulses can be used to spatially encode the MRI signal. An AFP pulse can be used to invertthe spins and leave them with an approximately linear B1-dependent spatial phase, ψ^y^,where y is a spatial dimension. In general, FREE imaging methods can be used to preciselycontrol ψ^y^ to spatially encode the spins or signal with a B1 field gradient duringtransmission of an RF pulse (e.g., AFP pulse). FIG. 1 shows a non-limiting example that5 QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582demonstrates phase accrual of different spins experiencing a B1 gradient duringtransmission of an AFP pulse. The ωeff trajectory, during the pulse, is shown for threedifferent example spatial locations, r1, r2, r3. Across a B1 field, the FM component (Δω(t))remains constant, while the amplitude modulated component (ω1(rn, t)) varies with theamplitude of the RF transmission field. If the adiabatic condition is held, the spins remainapproximately perpendicular to their respective ωeff fields and accrue a B1-dependentphase throughout the pulse, ψ(t). ψ(t) is B1-dependent, and thus spatially dependent, andcan be used for encoding in FREE pulse sequences.

[0041] As a non-limiting example, FREE methods can be applied to provide phaseencoding, as described in co-pending U.S. Patent Application No. 18 / 558,062, the entirecontents of which is incorporated herein by reference. For example, a bijective B1(y)transmits two AFPs in a double-spin echo pulse sequence with a difference of the time-bandwidth products (ΔR). ΔR creates a residual B1 phase that can be used for phase-encoding. ΔR can be stepped to produce phase encoding in a multi-shot or echo-trainapproach to cover multiple phase encoding lines of k-space. Such method may be referredto as PE-FREE.

[0042] The present disclosure describes FREE methods that achieve frequencyencoding, which may be referred to as FE-FREE. In some implementations, FE-FREE canbe used alone to achieve 1D spatial encoding. In other implementations, FE-FREE can beused in combination with PE-FREE or another spatial encoding technique (e.g., sliceselection, B0 encoding) to achieve 2D spatial encoding.

[0043] In FE-FREE, following an excitation pulse, a bijective B1(y) transmits twoAFPs sequentially. The first AFP can be used to dephase the spins, while the second AFPrefocuses them, forming an echo that can be measured. As a non-limiting example, FIG.2A illustrates an example of a pulse sequence 100 diagram that may be used to achieveFE-FREE. First, an excitation pulse 102 can be used to excite the MRI signal. The excitationpulse 102 may include a square pulse that excites the signal by 90° or another flip angle.The excitation pulse may also include another type of amplitude-modulated excitationpulse, such as a sinc, truncated sinc, windowed sinc, Shinnar-Le Roux (SLR), or others.The excitation pulse 102 may also include a frequency-modulated pulse, such as anadiabatic pulse, chirp pulse, or others. The excitation pulse 102 may be slice or slabselective RF pulse and can thus be played in conjunction with a slice or slab selectivegradient.6 QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582

[0044] After exciting the spins with an excitation pulse 102, a first frequency-modulated RF pulse 104 can be applied. The first frequency-modulated RF pulse 104 canhave an amplitude modulation 106 and frequency modulation 108 through time and bedescribed by a time-bandwidth product Ra. The first frequency-modulated RF pulse 104can be used to dephase the excited spins and may be referred to as the dephase (dp) pulse.Next, a second frequency-modulated RF pulse 114 can be applied to refocus, or rephase,the excited spins in order to form an echo that can be measured. Thus, the secondfrequency-modulated RF pulse 114 may be referred to as the readout (ro) pulse. Thesecond frequency-modulated RF pulse 114 can be described by an amplitude modulation116 function through time, a frequency modulation 118 function through time, and atime-bandwidth product Rb. The amplitude modulations 106 and 116 and the frequencymodulations 108 and 118 may be defined, for example, based on a standard HS pulse.However, the first frequency-modulated RF pulse 104 and the second frequency-modulated RF pulse 114 may alternatively be defined by other pulse types, such as chirp,BIR4, or others. In some implementations, the type of pulse used for the first frequency-modulated RF pulse 104 may be different than the type of pulse used for the secondfrequency-modulated RF pulse 114.

[0045] Both the first and second frequency-modulated RF pulses 104, 114 add andsubtract an amount of phase that is proportional to the time-bandwidth product, or R-value, of the respective pulse. Thus, the R-values can be chosen in order to create a rotaryecho during the second frequency-modulated RF pulse 114. For example, the firstfrequency-modulated RF pulse 104 and the second frequency-modulated RF pulse 114may have unequal time-bandwidth products, Ra and Rb, in order to create an echo duringthe second frequency-modulated RF pulse 114. As a non-limiting example, Rb may betwice that of Ra (Rb = 2Ra) in order to create an echo at the center of pulse B. Such arelationship between Ra and Rb can be achieved by adjusting the pulse length, the pulsebandwidth, or a combination of both. For example, to achieve Rb = 2Ra, the pulse durationof the second frequency-modulated RF pulse 114 (Tୠ୮) may be set to twice that of the firstfrequency-modulated RF pulse 104 (Tୟ୮), i.e., Tୠ୮ = 2Tୟ୮, while using an equal bandwidth.This condition creates a rotary echo that is centered at the second frequency-modulatedRF pulse 114 and can achieve a k-space trajectory that is similar to a gradient echo, orgradient recalled echo (GRE), pulse sequence.

[0046] In general, phase accrual achieved can be described as: 7 QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582t^(t ) ^^^eff^t ^^dt ^ , where t ^[0, T p ] . (4).0

[0047] ofy), in whicht^ (t , y ) ^^^eff ^t ^ , y^dt ^ , where t ^[0, T p ] and y ^[0, FOV y ] , (5)0

[0048] accrual104 and the second frequency-modulated RF pulse 114, which may be approximatelylinear in space, can be described by:ψ ^t, y^ ൌ ^ω ^tᇱ, y^dt′ , where t ∈ ୟୟ ^^ ^^^ ^0, T୮^ and y ∈ ^0, FOV^^ and (6).

[0049] echo that can be defined based on the phase accrual of the two pulses as:ψ୬^^^t, y^ ൌ ψୟ^t, y^ െ ψୠ^t, y^. (8).

[0050] frequency-modulated RF pulse 114, ψୠ^t, y^ can be set as twice that of ψୟ^t, y^ . Asdescribed above, one way to achieve this phase relationship is to double the pulseduration of the second frequency-modulated RF pulse 114 (Tୠ୮) compared to that of thefirst frequency-modulated RF pulse 104 ^T୮ୟ) while keeping the other pulse parametersconstant, for example.

[0051] Modifying the pulse duration (e.g., Tୠ ୟ୮ = 2T୮) provides a simple method forcreating an echo during the second frequency-modulated RF pulse 114, specificallycentered on the second frequency-modulated RF pulse 114. While this case is used hereinas an example, it should be understood that spatial encoding can be achieved in manyother ways. In general, ψ depends on ω୫ୟ^^ , Tp, A, β, and the pulse type. Thus, encodingcan occur through introducing pulse differences with any combination of these variables.For example, the two pulses may be equal in length where the second frequency-modulated RF pulse 114 has ω୫ୟ^^ twice that of the first frequency-modulated RF pulse104. Moreover, the relationship between the phase of the first and second phases( ψୟ^t, y^ and ψୠ^t, y^ ) need not be constrained to ψୠ^t, y^ ൌ 2ψୟ^t, y^ . Otherrelationships between ψୟ^t, y^ and ψୠ^t, y^ can be used, which may shift the echo from8 QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582the center of the pulse and provide asymmetric k-space sampling.

[0052] The signal equation during the AFP can then be described as: FOV s(t) ^^ Mxy(t,y)exp( ^i^ (t,y))dy , (9)

[0053]

[0054] The approximately linear phase and produced by thefirst frequency-modulated RF pulse 104 and the second frequency-modulated RF pulse114, respectively, can be used to fill the readout direction of k-space through time, t. Thek-space trajectory can be defined as:^ந ^k^t^ ൌ ^൫^౦ ,^൯െ ^ந^^^,^^, wh b^^ ^^ ere t ∈ ^0, T୮^. (10).

[0055] as:k୫ୟ^ ൌ ౦^^ െ ౦^^ , (11),

[0056] Δy ൌ 2k^^౮ (12),

[0057] FOV ൌ N ⋅ Δy ൌ N ^^ ro ro ^ଶ୩^^౮^, (13)

[0058] where Nrois the number of points acquired during the echo, spaced by ∆k.∆k, which is a function of time, is dependent on the pulse selection and sampling protocol,as will be described below.

[0059] Referring now to FIG. 2B, the pulse sequence diagram of FIG. 2A isexpanded to show an example sampling pattern that can be used to measure the echo 202formed by the combination of the dephase pulse (e.g., the first frequency-modulated RFpulse 104) and the readout pulse (e.g., the second frequency-modulated RF pulse 114).As previously described, an echo 202 can be formed during the readout pulse, or thesecond frequency-modulated RF pulse 114. For example, the echo 202 may be centeredon the second frequency-modulated RF pulse 114. In order to measure the echo 202, thesecond frequency-modulated RF pulse 114 may be intermittently applied. For example,the RF transmitter may be repeatedly turned on and off. This intermittent RFtransmission can create transmit periods 204 and gaps 206. During the transmit periods204, the transmitter can be turned on (i.e., “Tx on”), while the RF receiver can be turnedoff (i.e., “Rx off”). This transmit period 204 can have a length described by τ୮210. During9 QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582the gap periods 206, the transmitter can be turned off (i.e., “Tx off”) while the RF receivercan be turned on (i.e., “Rx on”) to measure the signal. In this way, the echo 202 can bemeasured nearly simultaneously with the second frequency-modulated RF pulse 114.

[0060] The gap period 206 can be further divided into two segments. The firstsegment may be a delay period of length τ୰^ 212, chosen to avoid ring-down effects of thereceive coil; the second segment may be a readout period of length τ୰୭214 during whichone or more k-space points may be measured. The delay in the first segment, τ୰^, may bechosen empirically based on the length of the ring-down effects. Data can be acquiredduring the full gap period 206, during only readout period (τ୰୭214), or some other subsetof the gap period 206 that includes τ୰୭214. If data are acquired during the ring-downperiods (τ୰^ 212), it can optionally be discarded during reconstruction. Data acquiredduring τ୰୭ can be used in the reconstruction to produce an image of the subject or object,which was spatially encoded by the frequency-modulated pulses. This gapping strategycan be used with minimal effect on pulse performance.

[0061] Thus, in some examples, delays of length ^^^ௗ ^+^ ^^^^ may be placed inbetween pulse elements ( ^^^) of the second AFP. ^^^ௗ may be inserted to allow coil ring-down, as shown in FIG. 2B, and data acquired during ^^^^can be used for reconstruction.Gapping minimally affects pulse performance and permits sampling of the rotary echo.

[0062] Referring now to FIG. 2C, another example pulse sequence is provided,which can be used to provide frequency encoding using B1 gradient fields. As previouslydescribed, the transmitter may be repeatedly turned on and off during the AFP pulse tosample during the pulse. Signal is acquired during the gaps of the second AFP pulse, whilethe first AFP pulse is gapped to maintain the refocusing condition (e.g., described byEquation 10). The pulse can be split into three segments, including the pulse segment(τ୮), the ring down period (τ୰^^, and readout period (τ୰୭). In this way, to reduce the effectof ringdown, the pulse segments can be defined according to the waveform x(t), which isa normalized version of Equation 2, that smoothly brings the transmitter amplitude tozero. Thus, the pulse may be described as ^^ ^^ஈ^ ^^^ ൌ ^^^^ ^^^^Π^^ ^^ െ ^^ ^^ ^^^ ∗ ^^^ ^^^^ whereΠ^ defines a comb function, ω^^ ^^^ is the gapped AMand ^^ ^^ ൌ ^^^ ^ ^^^ௗ ^ ^^^^ is thespacing of the pulse segments.

[0063] In this example, a rotary echo is created by setting the time-bandwidthproduct of Rro equal to twice that of Rdp. The rotary echo is centered during Rro and gapsof length ^^^ௗ ^ ^^^^ are inserted between pulse segments of length ^^^ . These pulse10QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582segments are defined by a waveform that smoothly brings the transmit amplitude to zeroto reduce ringdown. In some implementations, signal generated during ^^^ௗ is avoided(e.g., not measured or discarded) due to coil ringdown, and data acquired during ^^^^ isused to reconstruct an image. These pulses can be run sequentially, similar to thedephasing and refocusing lobes of a gradient recalled echo sequence (GRE), but the useof B0 gradients is not required.

[0064] In FE-FREE, signal may be sampled while the second AFP executes a 180-degree rotation of the transverse magnetization. In the first (frequency-modulated)rotating frame, the transverse magnetization lacks the conjugate relationship found in atypical echo. The rotary echo splits into Mx, My, and Mz magnetization components. In asecond reference frame rotating at the angular velocity defined by the effective fieldangle, α(t), the Mx and My contain a conjugate relationship, where α(t) is defined as:^AM ^^ ^t ^ ^arctan ^ t ^^FM (14).^ t ^^ ^ ^ ^ ^

[0065] To reach the second rotating frame, a y-rotation matrix using α(t) of thepeak γB1 may be applied to the data:^Sx ''( t )^ ^ cos ^ ^^^ t ^ ^ 0 sin ^ ^^^ t ^ ^ ^ ^Sx '( t )^^ ^ ^ ^ ^ ^;

[0066] inthe first and second rotating frames, respectively. The transverse signal (Sx ^ and Sy ^) maybe sampled through traditional approaches, and the longitudinal signal, Sz ^(t), may beestimated as:Sz ^ ^ t ^ = S max - S xy ^ ^ t ^ (16);

[0067] where Smax is the maximum of the transverse signal, Sxy ^(t). The object isreconstructed by Fourier transformation of Sx ^ ^ and Sy ^ ^.

[0068] As described above, FREE can use AFP pulses to produce spin echoesinsensitive to B0 and B1 inhomogeneity. The accumulated phases of the transversemagnetization during the AFPs are spatially dependent and described by ^ ^r, t ^ :t r^^ = ^ ^ ^ dτ wheret ^ T o^ r and ;;UMN 2022-135Quarles Ref.920171.00582

[0069] where t is a dummy integration variable; t is time; r is spatial distance; rmaxis the field-of-view (FOV); Tp is pulse length; AM(t) and FM(t) are the AFP’s pulseamplitude- and frequency- modulation functions; A = BW ^ / 2, where BW ^ is the pulseexcitation bandwidth; and w1maxI is the peak value of γB1 at position r. ψ(r,t) isapproximately linearly dependent on ω1I(r).

[0070] In an example implementation, FE-FREE uses of two successive AFPsfollowing an excitation pulse, as shown in FIG. 2B. The resulting spatial phase is definedby:^ dp ronet ^r ^ ^ ^ dp ^ r, T p ^ ^ ^ ro ^ r, T p ^ where r ^[0, r max ] (19);

[0071] ψro define the phase from the first and second AFP. Both pulses may be transmitted withthe same BW ^, FM(t), and B1 gradient. By setting Tpro = 2Tpdp, a rotary echo centered atthe middle of the second AFP can be created.

[0072] As described above, Hyperbolic secant (HS) pulses may be used for bothAFPs due to their smooth inversion profiles across B0 inhomogeneities. ∆k, defined as:^k( t ) ^ d ^ ro ( r , t ), wheret ^[0, Trop ] and r ^[0, rm] (20). drax

[0073] order HS pulses (e.g. HS10) present lower variations in ∆k. Thus, ∆k(t) depends on thepulse type. ∆k may not be equal during the whole readout train. However, the pulse typemay be chosen in order to reduce variation in ∆k(t) during acquisition. For example, FIG.3A illustrates ∆k as a function of time for HS pulses of the 1st (HS1), 5th (HS5), and 10th(HS10) order. In this non-limiting example, ∆k(t) was simulated for a linear γB1 gradientof 9 kHz ∙ (5 cm)-1. Δk(t) approximately varied according to the shape of the amplitudemodulation function. The higher-order HS pulses can provide more uniform k-spacesampling in the center regions of the pulse.

[0074] As another non-limiting example, FIG. 3B shows simulated k-spacetrajectories for HS1, HS2, HS3, HS4, and HS5 pulses used as a readout pulse in FE-FREE.In this example, the k-space trajectory was calculated using Equation 10. The first andsecond AFPs had a time-bandwidth product of Ra = 2000 Hz · 0.02 s and Rb = 2000 Hz ·0.04 s. The pulses were transmitted with gradient of 7 kHz ∙ (10 cm)-1 and RF peak ^^ ^^^ / 2 ^^of 10 kHz. Higher-order HS pulses covered more of k-space and had a larger linear Δkregion. Non-linearities in spatial frequencies can arise from spins’ nutation frequency,12QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582^^^^^^ ^^^. This behavior can be predicted due to its dependence on the pulse bandwidth.Selecting A values approximately equal to the weakest γB1(y) can help reduceand hold positions in space at a constant spatial frequency during periodsof acquisition, which may be advantageous, especially when using a Fourierreconstruction. The effective nutation frequency (ω^^^^t^) is plotted in FIG. 4A over thepulse duration for varying values of the peak Rabi frequency during the pulse (A). In thisnon-limiting example, ω^^^^t^ was simulated for an HS10 pulse with γB1 = 8 kHz for an Avalue of 1, 5, and 10 kHz. Effective nutation frequencies are plotted for a second non-limiting example in FIG. 4B. In this example, an HS10 pulse was used with Tp = 20 ms.^^^^^^ ^^^ / 2 ^^ was calculated for A / 2π = 1, 5, and 10 kHz and for ^^^^^௫ / 2 ^^ = 3, 5, and 7 kHz.

[0075] To allow for Fourier reconstruction, the effective spatial frequency shouldpreferably remain constant or nearly constant during acquisition. As demonstrated inFIGS.4A-4B, non-linearities in the spatial frequencies can arise based on the relationshipbetween γB1(y) and A. When A is much greater than any region of γB1(y), the spins reachtheir peak Rabi frequency of A and eventually settle at γB1(y). This settling time dependson the difference between γB1(y) and A, being longer for increasing values of A. Thus,setting A much greater than any value of γB1(y) (e.g., A = 10 kHz) may introducenonlinearities in ω^^^^t^ and ∆k(t), which degrades Fourier reconstruction. Selecting Avalues approximately equal to the weakest γB1(y) (e.g., A = 1 kHz) can avoidnonlinearities and hold each position in space at a constant spatial frequency duringperiods of acquisition. In some implementations, data acquired from higher bandwidthpulses may benefit from non-linear Fourier reconstructions due to the non-linear spatialencoding achieved.

[0076] Assuming a linear B1 field gradient, a small A value, and a constant Δk(t),the analytical form of Δk can be defined as:Δk ൌ ^நmaxିநmin^^^ , (21),

[0077] (y)field values,and Δy is the spatial extent of the coil.

[0078] The trajectory of spins produced by an FE-FREE pulse sequence (e.g., FIG.2A-2C) can be considered in a first rotating frame that follows the Larmor frequency, asdemonstrated in FIG. 5. The rotary echo can be sampled while a second frequency-modulated pulse (e.g., AFP) produces Beff causing a rotation ωeff (e.g., 180° excitation) of13QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582the transverse magnetization. In the first rotating frame, the rotary echo can be brokendown into x-, y-, and z-components. The ωeff field rotates through the x–z plane enablingthe y-component to remain detectable with conventional approaches. The echo’s x- andz- components depend on the inversion rate of the pulse, α^t^. A slower inversion resultsin more signal in the x-component of the spins, while a faster inversion results in moresignal in the z-component. For example, the magnetization components (Mx, My, and Mz)are plotted in FIG. 6 for an HS1 and HS10 pulse. The HS1 pulse, which achieves a slowerinversion, is characterized by larger signal in the x-component of the spins, while theHS10 pulse, which causes faster inversion, achieves larger signal in the z-component ofthe spins.

[0079] Fourier reconstruction can be considered in a second frame of referencerotating at the angular velocity defined by α^t^. As described above, the transverse signal(Sx ^(t) and Sy ^(t)) can be sampled using traditional approaches, and the longitudinal signal,Sz ^(t), by calculation:Sz ^ ^ t ^ = S max - S xy ^ ^ t ^ . (22).

[0080] can be transformed into the second reference frameusing a y-rotation matrix that is dependent on α^t^, such as the one shown in Eqn. (15).While S^ᇲ and S^ᇲ in Eqn. (15) lack a conjugate relationship, S^ᇲᇲand S^ᇲᇲ are conjugates ofeach other. Thus, the second rotating frame permits reconstruction with the Fouriertransformation.

[0081] FIG. 7 shows an example of magnetization components plotted withrespect to the first and second rotating frames. The spins shown have a 500 Hz offsetduring the second AFP of the FE-FREE pulse sequence (e.g., FIG. 2A). In the first rotatingframe, the signal is made up of Mx, My, and Mzmagnetization components. The Mxand Mylack a conjugate relationship, preventing the use of Fourier reconstruction. In the secondrotating frame, Mx and My remain conjugates of each other. The Fourier transformationcan be used in the second rotating frame in order to reconstruct the signal.

[0082] Referring now to FIG.8, a flowchart is provided that illustrates an exampleprocess 800 for using FE-FREE to achieve image acquisition. The process 800 begins inblock 802, during which a subject, patient, or other sample can be placed into an MRIsystem. Such system may include a permanent or superconducting main magnet toproduce a static magnetic field, called B0. The system may also include an RF transmitsystem and an RF receive system, that may be integrated together or separate systems.14QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582Advantageously, the system may not be required to include a B0 gradient system, as theRF transmit system can be used to generate a B1 field that can achieve spatial encoding.However, a B0 gradient system may be provided along one or more dimensions in someimplementations.

[0083] Next, an RF pulse can be used to excite a signal in block 804. As a non-limiting example, the pulse in block 804 may be a hard pulse that achieves 90° excitation,such as the RF excitation pulse 102 in the pulse sequence of FIGS. 2A-2C. The excitationmay be applied in conjunction with a slice or slab selective gradient in order to achieveslice or slab selection. The excitation may also include a spatially selective pulse used toachieve slice, slab, or another spatial selection, for example, without the use of anadditional gradient. The excitation may have any desired flip angle, such as 90°, oranother angle chosen to achieve a desired contrast.

[0084] Next, blocks 806 and 808 can be used to achieve spatial encoding using firstand second frequency-modulated RF pulses in accordance with the FE-FREE techniquesdescribed in the present disclosure. For example, an echo can be formed as the slope of alinear gradient varies through time. The echo can be measured by repeating block 814several times to sufficiently sample the echo. The echo formed by blocks 806 and 808 isachieved in a similar manner as a gradient echo sequence, where block 806 can beconsidered analogous to the dephasing gradient lobe, and block 808 can be thought of asthe rephasing gradient lobe that is twice as long as the dephasing gradient.

[0085] In block 806 a frequency-modulated RF pulse (e.g., the first frequency-modulated RF pulse 104) can be applied to dephase the signal. For example, thefrequency-modulated RF pulse may be a hyperbolic secant pulse of any order (e.g., HS1,HS5, HS10). As previously described, the pulse can be defined with an amplitude andfrequency modulation in order to supply a spatially varying phase to the signal. Forexample, the phase may be linear in space to have a similar effect on the signal as a linearspatial B0 gradient.

[0086] In block 808, the signal can be refocused and measured nearlysimultaneously. Block 808 can be applied to begin signal refocusing either immediatelyafter the dephasing pulse or after a delay, which can be chosen to achieve a desiredcontrast, for example.

[0087] To refocus, or rephase the signal, a frequency-modulated RF pulse (e.g., thesecond frequency-modulated RF pulse 114) can be applied in block 808. As previously15QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582described, the refocusing RF pulse can have an amplitude and frequency modulation,which can be chosen in order to refocus the signal and form an echo. For example, thepulse can be generated in order to create a linear, or approximately linear, phase acrossthe imaging field of view. By choosing the appropriate pulse parameters, the phase of therefocusing pulse may be twice that of the dephasing pulse, which can create an echocentered on the refocusing pulse.

[0088] In some examples, to measure the echo formed during the RF pulse therefocusing pulse can be applied intermittently, one segment at a time, in block 812. Forexample, block 812 may include the transmit period 204 of length τ୮210. After a segmentof the refocusing pulse is played in block 812, the pulse can be paused in block 814 inorder to measure a segment of the echo. For example, block 814 can include the gapperiod 206, in which a delay period (e.g., τ୰^ 212) can be used to reduce the effects ofring-down of the coil and the signal used for reconstruction can be measured during areadout period (e.g., τ୰୭214). The signal can be intermittently measured in block 814using an RF receive system.

[0089] Blocks 812 and 814 can be repeated N times in order to sufficiently samplethe echo. For example, N could be NRO, such that a full readout line of k-space is acquiredwith each refocusing pulse. If τ୮ is equal through every repetition 816, then N can berelated to the pulse duration (e.g., Tୠ୮) and the length (e.g., τ୮210) of the transmit period^^204 as τ୮ ൌ ౦^ , for example. In some other examples, blocks 812 and 814 may not beequalevery repetition 816. For example, τ୰^ may vary through repetitions 816,if desired.

[0090] As another non-limiting example, block 814 can be omitted in some of therepetitions of block 816. For example, the signal can be sampled in block 814 duringsegments of the pulse that produce a high degree of phase linearity. For example,referring back to the HS5 and HS10 pulses illustrated in FIG.3, it may be advantageous toomit sampling during the beginning and end pulse to sample the echo when Δk is mostconsistent. In the example HS10 pulse shown in FIGS. 3A, the sampling may be omittedfrom 0 to 5 ms and again from 20 ms to 25 ms to avoid variable k-space sampling, ifdesired. In other instances, another subset of the pulse or even the full pulse may be alsosampled in block 808, if desired.

[0091] Blocks 804, 806, and 808 may be repeated M times. For example, repeating 16QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582the FE-FREE pulse sequence can be used to acquire multiple averages of the readout lineor to fill out k-space in the phase encoding direction, in which case, phase encoding maybe added prior to dephasing the signal in block 806. For example, M could be equal to NPE,NAvg, or NPE ൈ NAvg. Additionally or alternatively, block 804 may be excluded in some or allof the repetitions 820 if the FE-FREE pulse sequence is being used as a part of a multi-echo sequence in which several phase encoding lines of k-space are measured after asingle excitation. The excitation of block 804 may also be excluded to repeatedly acquirea line of k-space with varying echo times to produce varying contrasts, such as T2mapping.

[0092] Other steps may be added to process 800. For example, phase encodinggradients or slice selective gradients may be added to achieve spatial encoding in otherdimensions. Phase encoding may also be achieved without the use of B0 gradients byadding PE-FREE encoding or other related techniques to the process 800. Othergradients, spoilers, B1 fields, dummy pulses, or pulse sequence gaps may be added inorder to achieve a desired contrast. As yet another non-limiting example, the process 800can be expanded to apply the FE-FREE technique along two or more spatial dimensions.For example, applying FE-FREE (e.g., blocks 806 and 808) synchronously on twoorthogonal coils could generate a radial k-space trajectory. Introducing longitudinalacquisition and radial imaging could permit fast pulse sequences.

[0093] Once k-space is sufficiently sampled (e.g., in block 808 through therepetitions at blocks 816 and 820), image reconstruction can be performed in block 810.The image reconstruction can use the data acquired during the block 814, which may belimited to the data acquired during the readout periods (e.g., τ୰୭214), discarding dataacquired during the ring-down periods (e.g., τ୰^ 212). The image reconstruction caninclude Fourier reconstruction, as previously described. In other examples, imagereconstruction can be implemented using an iterative reconstruction algorithm, amachine learning-based image reconstruction algorithm, or the like. The imagereconstruction may also include artifact correction or further image processing, such asgenerating quantitative maps, image segmentation, etc.

[0094] The image reconstruction of block 810 may also be informed by thesampling pattern. For example, the image reconstruction may require reconstruction ofnon-uniformly sampled k-space (e.g., radial). Thus, gridding techniques may be used, aswell as other known methods, such as model-based reconstruction, iterative17QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582reconstruction, machine learning techniques, and others.

[0095] As will be described in further detail below, merging FE- and PE-FREE canprovide for 2D or full volumetric imaging without dedicated B0 phase or frequencyencoding gradient coils. The merging of FE- and PE-FREE can enable advanced B1 pulsesequences that are similar to EPI, PROPELLOR, or steady-state imaging.

[0096] In some implementations, 2D encoding can be achieved using acombination of PE-FREE and FE-FREE, which may be referred to as a 2D FREE pulsesequence. In 2D FREE, PE-FREE and FE-FREE can be applied using separate B1 fieldgradients. These B1 field gradients may be perpendicular to one another to achieve 2Dencoding along two orthogonal directions. In this way, multi-shot 2D FREE can achieve asimple 2D cartesian k-space trajectory that can be reconstructed using a standard 2DFourier transform or other reconstruction techniques. In other implementations, otherk-space trajectories can be used.

[0097] The combination of PE- and FE-FREE along two dimensions is able toincrease acquisition time for FREE imaging, advantageously making the use of FREEtechniques clinically feasible. Moreover, the use of B1 gradients to achieve both frequencyand phase encoding can further reduce MRI costs by eliminating the need for B0 encodinggradient systems.

[0098] FIG. 9 provides a non-limiting example of a 2D FREE pulse sequencediagram in which PE-FREE and FE-FREE are combined to achieve spatial encoding alongtwo directions (e.g., x and y). In this example, PE-FREE provides spatial encoding alongthe y direction, and FE-FREE provides spatial encoding along the x direction. Anexcitation pulse is applied in the presence of a slice selection gradient to excite a slicealong z. Phase encoding is provided by generating a pair of frequency-modulated RFpulses, labeled Ra and Rb. Next, frequency encoding is provided using a pair of gappedfrequency-modulated pulses, labeled Rdp and Rro. An echo is formed in the center of Rro,filling a readout line of k-space as measured by the analog-to-digital converter (ADC).This pulse sequence can be repeated to fill in k-space in the phase encoding direction andcan be repeated for varying slice selection gradients to provide imaging of a volume ofinterest.

[0099] Referring now to FIGS.10-11, simulation data are presented in accordancewith some aspects of the present disclosure. During FE-FREE, residual z‐magnetizationlocks to ^eff during its trajectory. Residual magnetization could arise from unexcited18QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582magnetization by imperfect 90-degree pulses (FIG. 10) or other pulse imperfections.Spin-locking produces a signal that mirrors the pulse profile. The data can be correlatedto remove the pulse waveform, which, in some implementations, is approximately equalto the spin-locked signal:ோ^ ∗^^^ ^^^ ൌ ఠ^^ ^ఠ^|^^ఠ^|మ (23)

[0100] thespin-densityusing Equation 23.

[0101] Equation 23 assumes that the spin-locked component matches the pulsewaveform; an assumption named the spin-locked matching approximation (SMA). Thisapproximation enables the removal of most spin-locked signal from the data (FIG. 10).Divergences from SMA result in residual magnetization being present post correlation(FIG.10) and can arise from a loss of adiabaticity or variable ^eff(t) trajectories.

[0102] To study the accuracy of SMA, a propagator analysis was performed. Apropagator matrix (Equation 24) was defined as follows, which inverts spins with perfectadiabaticity.

[0103] ^ ^during an AFP pulse. A simulation was performed to study the evolution of spin-lockedcomponent in different conditions. FIG. 11 shows the effect of keeping the pulse lengththe same but changing the bandwidth. Under low bandwidth conditions, the discrepancybetween the normalized spin-locked real component and pulse AM profile is minimal.However, as the pulse bandwidth increases, the ^eff(t) trajectory, in the FM frame,deviates from the pulse profile. Like the condition observed in FIG. 4B, a low bandwidthpulse allows SMA to remain valid, facilitating the recovery of the desired data throughcorrelation. Conversely, larger bandwidths reduce the accuracy of the SMA and Equation24 in isolating the rotary echo from the spin-locked components.

[0104] Referring again to FIG. 10, a simulated rotary echo created by FE-FREE isprovided. The rotary echo’s magnitude, real, and imaginary components are plotted fortwo cases: a perfect 90-degree excitation across space (column A) and an imperfectspatial excitation produced by a B1 gradient (column B). A 50 μs hard pulse excited the19QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582spins in both cases. In the uniform excitation case (left column), a γB1 / 2π set to 5 kHzachieved a uniform 90-degree flip across space, while in the B1 gradient case (rightcolumn) a B1 gradient with a minimum and maximum value of 0 and 5 kHz excited spins.The data is acquired in the phase-modulated frame of the pulse. The simulated object wasa boxcar centered in space that covered 50% of the FOV. In both cases, spin-lockedcomponents are seen that track the waveform of the pulse, but a uniform excitationdrastically reduced spin-locking. Correlation was applied to both cases. The bottom rowshows the result post-correlation. Correlation extracted the rotary echoes of interest ineither case, but excitation with a B1 field gradient reduced the available signal to beextracted.

[0105] FIG. 11 shows an adiabatic full passage simulated in the FM frame using apropagator analysis defined in Equation 24. The pulse was simulated to invert spinsaligned with ’he ^z' axis. The B1 gradient was 5 kHz / 5 cm, and a peak amplitude oftransmitted pulse was γB1 / 2π ൌ 7 kHz. Various bandwidths were considered. In all cases,an HS10 pulse with a duration of 20 ms was utilized. By plotting the real component ofthe tracked spins and the corresponding normalized AM^t^ profile, a discrepancybetween the correlated signal and true signal is shown. It can thus be observed that atlower bandwidth values, the pulse waveform accurately defines the trajectory of the spinlocked component, which correlation would remove.

[0106] In another example study, FE-FREE was tested for 1D imaging using a 10mL NMR tube phantom placed in the center axis of a 5-cm surface coil. In the phantom, aTeflon plug (width = 5 mm) separated 2 water chambers, as shown in FIG. 12.Experiments were performed on a custom 0.5 T Halbach magnet with ± 10 kHz of B0 off-resonance across the object. The 0.5 T system did not include any B0 gradient coils butwas situated with a B1 surface coil.

[0107] The surface coil was used to transmit a non-linear γB1gradient of ~10 kHz∙ (5 cm)-1. The pulse sequence was written in a custom single Tx / Rx CIERMag DigitalMagnetic Resonance Spectrometer operating at 21.1 MHz. A 100 μs square pulse wasused to excite the spins. The first and second HS10 time-bandwidth products were 0.020s ∙ 2,000 Hz and 0.040 s ∙ 2,000 Hz, respectively. τ୮ ൌ 80 μs, τ୰^ ൌ 50 μs, and τ୰୭ ൌ 5 μs.1-ms delays were inserted before and after the first HS10 pulse to allow the B0inhomogeneity to spoil the transverse signal. One point was acquired per τ୰୭, and 450total points were reconstructed. α^t^ was defined using a γB1 of 10 kHz.20QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582

[0108] As shown in FIG.12, the experiment demonstrated the ability to produce arotary echo in a highly inhomogeneous magnet without the use of B0 gradient coils. Suchan echo provided a good reconstruction of the 1D object.

[0109] Experiments were performed on a 0.5 T Halbach magnet with ±8 kHz of B0inhomogeneity across the object. The magnet has a weight of 150 kg, bore of 19 cm, andlength of 40 cm. The 0.5 T system has no passive shims, faraday cage, filter panel, or B0gradient coils. A heat regulator system, designed by RE magnet studio, kept the magnet’stemperature steady at 28 °C, reducing field drift, and a grounding panel reduced RF noise.The system uses an active Varian T / R switch, preamplifier (model: 0191432205), andCPC 500 W amplifier (model: 9T500M-16C). A two channel Tx / Rx CIERMag DigitalMagnetic Resonance Spectrometer (DMRS) operating at 21.1 MHz deployed the pulsesequence. The receive system has 16 bits of resolution, collecting one data point permicrosecond. Th’ DMRS's transmit channels can amplitude modulate in a 4-quadrantapproach with 16-bit amplitude dynamics, as well as frequency (24 bits with 0.373 Hzresolution) and phase modulation (14 bits with 0.022° resolution). A Python MRframework (PyMR) controlled the spectrometer and the system was equipped with anintegrated development environment (IDE) specialized for MR sequence design, and thecode was deployed in F-language.

[0110] For FE-FREE, a 10 cm surface coil transmitted a non-linear γB1 gradientwith an average value of ~2 kHz ∙ (5 cm)-1. A 150 μs square pulse excited the sample. Thesquare pulse was calibrated to produce an approximate 90-degree flip at the front of thephantom. HS10 pulses were utilized for both inversion pulses; the first and second time-bandwidth products were 0.010 s ∙ 4,000 Hz and 0.020 s ∙ 4,000 Hz, respectively, with τp= 50 μs, τrd = 5 μs, and τro = 45 μs. The parent pulse was subdivided by pulse segmentsdefined by twenty points (of length τp), and twenty points were acquired during τrd + τroby using an Rx bandwidth of ±200 kHz. The pulse segments (e.g., x(t)) were defined withthe AM(t) of an HS10 pulse. Data acquired during τrd were zeroed out. The data wascorrelated with the pulse waveform then down sampled by averaging to one point pergap, resulting in 200 points. The object was reconstructed by applying the Fouriertransform to the middle 64 points of the data. A three-chamber saline phantom wasimaged (FIG.13).

[0111] For PE-FREE, an optimized multi-shot DSE PE-FREE pulse sequenceincluded phase encoding with the following parameters: Rmin = 88, Rmax = 200, ΔR = 3.5,21QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582and all pulses had a BW = 10 kHz. The same excitation pulse as FE-FREE was used withthe same power settings, and TE = 55 ms, TR = 9 s, and 64 points were acquired per echowith an Rx bandwidth of ±15 kHz, resulting in 128 pts per echo. Five averages wereperformed per shot. The 2D Fourier transformation was applied to the N×M dataset,where N is the number of phase encodes, and M the number of points acquired per echo.The reconstruction shows the spin-density profile as a function of space and chemicalshift. The same three-chamber saline phantom was imaged for comparison (FIG. 13) toFE-FREE. In FIG. 13, the left panel shows a diagram of the 3-chamber phantom, B0 field,and B1 coil. The three chambers were filled with saline. The right channel shows a contourplot of a normalized Biot-Savart simulation demonstrating the approximate B1 field mapsand their relative location to the 3-chamber phantom.

[0112] FIG.14 shows simulation results for a 1D FE-FREE experiment. Spins wereassumed to begin on the transverse plane. A linear γB1 / 2π gradient of 12 kHz transmittedthe first and second AFP. The first and second AFP had a time-bandwidth product of0.0125 s ∙ 2000 Hz and 0.0250 s ∙ 2000 Hz, respectively. 800 points were collectedthroughout the second pulse. The complex Fourier Transform (FT) reconstructed themiddle 400 points. Due to the missing z-magnetization, the complex FT produced twocopies of the object about the zeroth frequency. The figure shows the positive frequencycopy across different off-resonant conditions. The object reconstruction begins toseverely degrade once the resonance is outside the A of the pulse. Simulation resultsshowed the ideal settings of a uniform excitation, linear γB1 field, and no RF ringdown.FE-FREE accurately reconstructed the middle slice of the Shepp-Logan phantom, asshown in FIG. 14. A constant resonance offset was simulated across space. Thereconstruction quality of FE-FREE is tolerant to resonance offset values that do notexceed the’pulse's bandwidth. However, as the resonance offset approached andexceeded the bandwidth, the reconstruction quality deteriorated, as shown in FIG. 14.The rotary echo occurs during the inversion of a plane. As a result, acquisition of only thetransverse signal fails to acquire the whole magnetization trajectory. Z-magnetization,which describes some of rotary echo, is lost. As a result, the complex FT distributes energyamongst the positive and negative frequencies symmetrically around the zerothfrequency. Only one copy was shown in the results.

[0113] FIG. 15 shows PE-FREE data reconstructed using a 2D FourierTransformation. A 2D contour plot (left) shows the three chambers and the B1 and B022QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582inhomogeneities present in the chambers. A 1D projection was generated (right) tocompare against FE-FREE (FIG. 16) by summing the rows of the 2D data. ExperimentalPE-FREE shows all three chambers of the object and the off-resonance distribution ineach object due to B0 inhomogeneity.

[0114] FIG. 16 shows the reconstruction of the FE-FREE rotary echo shows threechambers. The Fourier transform reconstructed the middle 64 points of the data.Distortions are present likely due to higher BW pulses and poor performance of theFourier reconstruction. Experimental FE-FREE images show a similar profile in the ω1dimension.

[0115] The present disclosure describes a novel methodology for frequencyencoding using B1 field gradients and AFPs. One goal of radiofrequency imagingapproaches is to offer alternatives to B0 gradient coils in novel MRI system designs. MostRF imaging approaches focus on phase-encoding, which images slowly and thus limitsimaging applications, or performs poorly in off-resonance conditions, which limitsapplications in compact magnets. In some aspects, both FE-FREE and PE-FREE canleverage adiabatic pulses to control the effective magnetic field to ensure uniform spinrotation across variations in B1 and B0.

[0116] The FE-FREE techniques described in the present disclosure are capable offrequency-encoding without dedicated B0 gradient coils. As described above, adopting agapping technique enables the use of delays between pulse elements to avoid ring-downartifacts and to allow for sampling rotary echoes. Adiabatic pulses can be used to give FE-FREE resilience to B0 off-resonance up to the bandwidth of the pulse. In one exampleprovided, 1D imaging was performed on a highly inhomogeneous (±10 kHz) 0.5 THalbach permanent magnet with no B0 gradient coils.

[0117] FE-FREE creates rotary echoes by transmitting AFP pulses with a B1gradient. As one non-limiting example, the AFP pulses may be generated using a 5 cmsurface coil to transmit the pulses. Alternatively, other RF coil geometries andconfigurations may be used. Like the dephasing lobe in a GRE, the first AFP pulsedephases the spins, while the second AFP pulse, like a refocusing lobe, refocuses the spins.Setting the pulse length of the second AFP pulse equal to twice that of the first, FE-FREEwas observed to create a rotary echo centered at the second AFP pulse. The rotary echoesoccurred during the inversion of a plane. While the first reference frame lacks a conjugaterelationship between transverse components, a second rotating frame set to the angular23QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582velocity defined by ^ ^t ^ can retain the conjugate relationship between the transversecomponents to permit a Fourier reconstruction.

[0118] In some implementations, ring-down delays, ^ rd , take up most of the gaptime. Optimized hardware or post-processing that minimizes ^ rd can significantlydiminish total pulse sequence length. Non-linearities in FE-FREE’s k-space trajectory maylimit sampling to the constant ∆k regions. In these instances, advanced reconstructionapproaches such as gridding, non-uniform Fourier transformations, or model-reconstructions can permit sampling and reconstructing of the nonlinear regions.

[0119] Traditional RF approaches detect the transverse component of FE-FREE’srotary echo. In the first rotating frame, the transverse component accounts for half of therotary echo. Most of its counterpart remains in the z-direction (as shown in FIG. 7). Theuse of longitudinal detection permits the acquisition of the Mz component andsimultaneous transmit and receive. In an example implementation, a longitudinal coildesign can be configured to detect low-frequency nutation, such as less than 15 kHz. Insome implementations, simultaneous transmit and receive can be implemented usingsignal detection within the Rabi frequency range, which is on the order of kHz. Thelongitudinal data can be merged with the transverse component to reconstruct the entiresignal. Alternatively, a similar reconstruction approach that solely uses the longitudinalcomponent can remove the need for gaps, thereby accelerating FE-FREE.

[0120] In an example study described above, the surface coil used for imaging hada B1 field length of ~5 cm. The development of optimized RF hardware with gradientsthat extend over the entire body can expand upon FE-FREE’s imaging capabilities.

[0121] FE-FREE approaches may be applied to any field strength. Low fieldstrengths may advantageously reduce limitations due to specific-absorption rate (SAR).To further minimize SAR, advanced coil designs and pulse sequence modifications thatdistribute SAR efficiently may be employed. Advantageously, 0.5 T can offer enough SNRto produce quality imaging with FE-FREE while simultaneously minimizing SAR.

[0122] The contrast in FE-FREE is dominated by T2ρ. 1 / T2ρ (= ½[1 / T1 + 1 / T2])contrast provides unique benefits to detecting slow metabolic functions and is muchlonger than T2 (T2ρ>>T2). The length of T2ρ enables the acquisition of many rotary echotrains. Taking advantage of T2ρ, merging PE-FREE, FE-FREE, and fast-spin echoapproaches may enable one-shot imaging.24QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582

[0123] With FE-FREE shown in principle, advanced k-space trajectories could bedeveloped that utilize the foundation of FE-FREE. One example is FE-FREE radial imaging.Applying FE-FREE synchronously on two orthogonal coils can generate a radialtrajectory. Introducing longitudinal acquisition and radial imaging can permit fast pulsesequences.

[0124] The experiments described provided one-dimensional images of aphantom in a highly inhomogeneous, compact 0.5 T Halbach system; the magnet had amaximum of 12 kHz 1H linewidth across the phantom. The magnet had no passive shims,B0 gradient coils, or Faraday cage. The inhomogeneity of the magnet would limit mostimaging approaches. However, FREE principles enabled acquisition of 1D projectionsusing both PE- and FE-FREE using a simple surface coil. These results, although one-dimensional, show the capability of FREE approaches to image with non-traditional MRImagnet designs. Expanding the abilities of FREE through advanced hardware and pulsesequences can promote the further use of magnets like the highly affordable, yetinhomogeneous 0.5 T magnet used in the provided examples. Merging FE- and PE-FREEand using the magnet’s natural B0 inhomogeneity to slice-select can enable volumetricimaging. A compact magnet powered by FREE, with no B0 gradient coils, can offer clinicalsolutions at a much lower cost than traditional systems.

[0125] Conceptually, FE-FREE shares similarities with traditional GREs. Typically,the dephasing lobe of a GRE brings spins to a -kmax, while the refocusing lobe sweepsthrough k-space to +kmax. FE-FREE can be used to mirror this k-space trajectory with thefirst and second AFP acting as the dephasing and refocusing lobes, respectively. Likewise,the area of the lobes and AFPs determines the resulting k-values. Similarities betweenFE-FREE and GRE approaches could enable traditional GRE post-processing tools to beeasily translated to FE-FREE to improve reconstructions, sampling, and trajectories.

[0126] In some implementations, a portion of the gap time is occupied by ring-down delays. Longer ring-down periods may be used when the system has poor coilloading, high Q-values, or lower magnetic field strength. The duration of the ringdownmay preferably be chosen as inversely proportional to the strength of the magnet. FE-FREE imaging can be used with mid to low field scanners, where RF coils dissipate energyover an extended period. Hardware (e.g., Q-switching) or post-processing techniques canbe optimized to ensure the utilization of all acquired data and minimize or reduce τrd.Thus, in some implementations a ringdown period may not be required.25QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582

[0127] In some example data, non-linearities in FE-FREE’s k-space trajectoryprevented the Fourier transformation from robustly reconstructing all data. However,modern reconstruction approaches can be used to reconstruct a nonlinear k-spacetrajectory (e.g., gridding, ramp sampling, non-uniform Fourier transformations, ormodel-reconstructions).

[0128] In some implementations, where A > ^B1, nonlinearities can arise in thespins’ spatial frequencies. Spatial-frequency variation during the pulse varies acrosspulse types (e.g., FIG. 4B) and may have a greater hinderance on reconstructionapproaches. However, to reconstruct these nonlinearities, a more detailed or advancedreconstruction can be used. In other implementations, the B1 gradient can be defined tohold the condition of ^B1 >> A, which removes spatial frequency variation. However,higher ^B1 fields could introduce higher levels of SAR.

[0129] Unlike a spin-echo,^ FE-FREE’s rotary echo uniquely refocuses in’t’e y'z'plane. Dependent on the pulse selection, adiabatic condition, and strength of the B1gradient, the rotary echo’has ’'-, y'-,’and z'- components. Traditional RF approaches detectthe transverse components but miss the longitudinal component. Not detecting thelongitudinal component can reduce SNR and produce a copy of the object. The object copyarises from the acqu’red x'’and y' component lacking the conjugate relationship that thecomplex FT requires. If the data is transformed into a second rotating frame, inverting ata specific angular velocity,’’he x''’’nd y'' hold the appropriate relationship and a rotaryecho becomes equivalent to a spin-echo in the first rotating frame. To appropriatelytransform into the second rotating frame, a calculation or measurement of the Mz(t)component can be used. In the z-direction, only the spins’ nutation is available, which isin the kHz range. Thus, longitudinal detection can be performed using low-frequency RFsensors that can measure the Rabi frequencies.

[0130] In the examples provided herein, the surface coil provided effective FE-FREE imaging, but it introduced spatial nonlinearities, residual Mz components, and alimited FOV. Typically, surface coils transmit a nonlinear B1 gradient that spatiallydistorts the image and varies the excitation flip angle. Distortion-correction algorithmshave shown to be effective at correcting surface coil distortions in FREE, but the spatialvariation of flip angles introduced large amounts of residual Mz that spin-locked to theAFP’s Beff trajectory. Magnetization that is spin-locked may not spatially encode.Correlation can be used to remove the prominent spin-locked component but filters out26QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582some signal. Advanced RF hardware that can transmit in a homogenous mode (e.g., a bird-cage) can be used to minimize spatial variations in flip angle and improve SNR. In theprovided examples, the 10 cm surface coil had an approximate maximum FOV of 10 cm.Spins outside of the coils reach were not imaged. However, coil designs for specificanatomy or that can image the whole body can be used to expand the FOV available forFREE imaging. Moreover, RF coils can be improved and designed specifically for B1encoding to provide highly linear B1 gradients, improving many aspects of FREE imaging.

[0131] In some implementations, FE-FREE uses minimal delays between itsdephasing and refocusing pulse; as a result, it measures the rotary echo. The rotary echo’sTE = Tpexc + 2Tpdp , where Tpexc is the pulse length of the excitation pulse. The refocusedspins relax in the presence RF field – otherwise known as T2 ^ ^relaxation. T2 ^ ^depends onthe RF pulse and spins’ overall environment. As applied, the spins T2 ^ ^vary across spacedue to the B1gradient, which may expand FE-FREE sensitivity toT2 ^ ^contrast ^ ^T2 ^ ^contrast provides unique benefits to detecting slow metabolic functionsand is much longer than T2 ^ ^The length of T2 ^ ^could enable the acquisition of many rotaryechoes during one experiment, potentially enabling one-shot imaging. Additionalcontrasts, like T1 and T2, can be introduced by varying the delays between the pulses,refocusing the spin-echo, which is produced by the first AFP, or varying the TR.

[0132] Typically, B1 imaging introduces more SAR than B0 approaches. SAR is B02-dependent. The MRI field has seen a recent rise in low to mid-field imaging technology,and its clinical applications are being thoroughly investigated. B1 imaging isadvantageously compatible with low or mid-field imaging. For example, lower magneticfields would allow simple, yet accurate, B1 mapping simulations, through the Biot-Savartequation, and allow FREE to bring its full capabilities to a scanner.

[0133] In recent years, there has been a surge of interest in the development ofaffordable, accessible, and portable MRI scanners. The re-emergence of ultra-low andlow-field scanners has gained significant attention and shows no signs of slowing downin their development. Among these innovations, B1 imaging approaches, such as FREE,present themselves as a potent addition to low-cost MRI systems. B1 imaging approachesmay provide an effective replacement of B0 gradient coils, especially for a low-fieldsystem. Such a system would hold immense potential for the ~90% of the world whereMRI remains inaccessible.

[0134] Referring particularly now to FIG.17, an example of an MRI system 1700 27QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582that can implement the methods described herein is illustrated. The MRI system 1700includes an operator workstation 1702 that may include a display 1704, one or moreinput devices 1706 (e.g., a keyboard, a mouse), and a processor 1708. The processor 1708may include a commercially available programmable machine running a commerciallyavailable operating system. The operator workstation 1702 provides an operatorinterface that facilitates entering scan parameters into the MRI system 1700. Theoperator workstation 1702 may be coupled to different servers, including, for example, apulse sequence server 1710, a data acquisition server 1712, a data processing server1714, and a data store server 1716. The operator workstation 1702 and the servers 1710,1712, 1714, and 1716 may be connected via a communication system 1740, which mayinclude wired or wireless network connections.

[0135] The MRI system 1700 also includes a magnet assembly 1724 that includesa polarizing magnet 1726, which may be a low-field magnet. The MRI system 1700 mayoptionally include a whole-body RF coil 1728 and a gradient system 1718 that controls agradient coil assembly 1722.

[0136] The pulse sequence server 1710 functions in response to instructionsprovided by the operator workstation 1702 to operate a gradient system 1718 and aradiofrequency (“RF”) system 1720. Gradient waveforms for performing a prescribedscan are produced and applied to the gradient system 1718, which then excited gradientcoils in an assembly 1722 to produce the magnetic field gradients in one or moredirections (e.g., G^ , G^ , and G^ ) that can be used for spatially encoding magneticresonance signals. The gradient coil assembly 1722 forms part of a magnet assembly1724 that includes a polarizing magnet 1726 and a whole-body RF coil 1728. However,with use of FE-FREE methods, the Gradient system 1718 and gradient coil assembly 1722may not be required and may be excluded from the system 1700. Alternatively, thegradient coil assembly 1722 may not be required to produce gradients in all threedimensions.

[0137] RF waveforms are applied by the RF system 1720 to the RF coil 1728, or aseparate local coil to perform the prescribed magnetic resonance pulse sequence.Responsive magnetic resonance signals detected by the RF coil 1728, or a separate localcoil, are received by the RF system 1720. The responsive magnetic resonance signals maybe amplified, demodulated, filtered, and digitized under direction of commands producedby the pulse sequence server 1710. The RF system 1720 includes an RF transmitter for28QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582producing a wide variety of RF pulses used in MRI pulse sequences. The RF transmitteris responsive to the prescribed scan and direction from the pulse sequence server 1710to produce RF pulses of the desired frequency, phase, and pulse amplitude waveform. Thegenerated RF pulses may be applied to the whole-body RF coil 1728 or to one or morelocal coils or coil arrays 1729. For example, the whole-body RF coil 1728 may be atransmit-receive coil or a transmit-only coil.

[0138] The RF system 1720 also includes one or more RF receiver channels. An RFreceiver channel includes an RF preamplifier that amplifies the magnetic resonancesignal received by the coil. For example, the whole-body RF coil 1728 can be connectedto a detector that detects and digitizes the I and Q quadrature components of the receivedmagnetic resonance signal. The magnitude of the received magnetic resonance signalmay, therefore, be determined at a sampled point by the square root of the sum of thesquares of the I and Q components:M ൌ ^^Iଶ ^ Qଶ^

[0139] and the phase of the received magnetic resonance signal may also bedetermined according to the following relationship:ϕ ൌ tanି^൬QI^

[0140] Other RF receivesuch as receive-only coils, local surfacecoils, multi-coil arrays, and other receive coils. The signal measured by multi-channel coilarrays can be combined in reconstruction using root sum of squares combination, parallelimaging methods, iterative reconstructions, machine-learning techniques, or otheradvanced methods, etc.

[0141] The pulse sequence server 1710 may receive patient data from aphysiological acquisition controller 1730. By way of example, the physiologicalacquisition controller 1730 may receive signals from a number of different sensorsconnected to the patient, including electrocardiograph (“ECG”) signals from electrodes,or respiratory signals from a respiratory bellows or other respiratory monitoringdevices. These signals may be used by the pulse sequence server 1710 to synchronize, or“gate,” the performance of the scan with the subject’s heartbeat or respiration.

[0142] The pulse sequence server 1710 may also connect to a scan room interfacecircuit 1732 that receives signals from various sensors associated with the condition ofthe patient and the magnet system. Through the scan room interface circuit 1732, a29QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582patient positioning system 1734 can receive commands to move the patient to desiredpositions during the scan.

[0143] The digitized magnetic resonance signal samples produced by the RFsystem 1720 are received by the data acquisition server 1712. The data acquisition server1712 operates in response to instructions downloaded from the operator workstation1702 to receive the real-time magnetic resonance data and provide buffer storage, so thatdata are not lost by data overrun. In some scans, the data acquisition server 1712 passesthe acquired magnetic resonance data to the data processor server 1714. In scans thatrequire information derived from acquired magnetic resonance data to control thefurther performance of the scan, the data acquisition server 1712 may be programmedto produce such information and convey it to the pulse sequence server 1710. Forexample, during pre-scans, magnetic resonance data may be acquired and used tocalibrate the pulse sequence performed by the pulse sequence server 1710. As anotherexample, navigator signals may be acquired and used to adjust the operating parametersof the RF system 1720 or the gradient system 1718, or to control the view order in whichk-space is sampled. In still another example, the data acquisition server 1712 may alsoprocess magnetic resonance signals used to detect the arrival of a contrast agent in amagnetic resonance angiography (“MRA”) scan. For example, the data acquisition server1712 may acquire magnetic resonance data and processes it in real-time to produceinformation that is used to control the scan.

[0144] The data processing server 1714 receives magnetic resonance data fromthe data acquisition server 1712 and processes the magnetic resonance data inaccordance with instructions provided by the operator workstation 1702. Suchprocessing may include, for example, reconstructing two-dimensional or three-dimensional images by performing a Fourier transformation of raw k-space data,performing other image reconstruction algorithms (e.g., iterative or backprojectionreconstruction algorithms), applying filters to raw k-space data or to reconstructedimages, generating functional magnetic resonance images, or calculating motion or flowimages.

[0145] Images reconstructed by the data processing server 1714 are conveyedback to the operator workstation 1702 for storage. Real-time images may be stored in adata base memory cache, from which they may be output to operator display 1702 or adisplay 1736. Batch mode images or selected real time images may be stored in a host30QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582database on disc storage 1738. When such images have been reconstructed andtransferred to storage, the data processing server 1714 may notify the data store server1716 on the operator workstation 1702. The operator workstation 1702 may be used byan operator to archive the images, produce films, or send the images via a network toother facilities.

[0146] The MRI system 1700 may also include one or more networkedworkstations 1742. For example, a networked workstation 1742 may include a display1744, one or more input devices 1746 (e.g., a keyboard, a mouse), and a processor 1748.The networked workstation 1742 may be located within the same facility as the operatorworkstation 1702, or in a different facility, such as a different healthcare institution orclinic.

[0147] The networked workstation 1742 may gain remote access to the dataprocessing server 1714 or data store server 1716 via the communication system 1740.Accordingly, multiple networked workstations 1742 may have access to the dataprocessing server 1714 and the data store server 1716. In this manner, magneticresonance data, reconstructed images, or other data may be exchanged between the dataprocessing server 1714 or the data store server 1716 and the networked workstations1742, such that the data or images may be remotely processed by a networkedworkstation 1742.

[0148] Referring now to FIG.18, an example image reconstruction system 1800 isshown, which may be used in accordance with some aspects of the systems and methodsdescribed in the present disclosure. As shown in FIG. 18, a computing device 1850 canreceive one or more types of data (e.g., signal evolution data, k-space data, receiver coilsensitivity data) from data source 1802. In some configurations, computing device 1850can execute at least a portion of an image reconstruction system 1804 to reconstructimages from magnetic resonance data (e.g., k-space data) acquired using an FE-FREEtechnique. In some configurations, the image reconstruction system 1804 can implementan automated pipeline to provide MRI images, quantitative maps, MRI data, etc.

[0149] Additionally or alternatively, in some configurations, the computing device1850 can communicate information about data received from the data source 1802 to aserver 1852 over a communication network 1854, which can execute at least a portion ofthe image reconstruction system 1804. In such configurations, the server 1852 can returninformation to the computing device 1850 (and / or any other suitable computing device)31QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582indicative of an output of the image reconstruction system 1804.

[0150] In some configurations, computing device 1850 and / or server 1852 can beany suitable computing device or combination of devices, such as a desktop computer, alaptop computer, a smartphone, a tablet computer, a wearable computer, a servercomputer, a virtual machine being executed by a physical computing device, and so on.The computing device 1850 and / or server 1852 can also reconstruct images from thedata.

[0151] In some configurations, data source 1802 can be any suitable source of data(e.g., measurement data, images reconstructed from measurement data, processed imagedata), such as an MRI system, another computing device (e.g., a server storingmeasurement data, images reconstructed from measurement data, processed imagedata), and so on. In some configurations, data source 1802 can be local to computingdevice 1850. For example, data source 1802 can be incorporated with computing device1850 (e.g., computing device 1850 can be configured as part of a device for measuring,recording, estimating, acquiring, or otherwise collecting or storing data). As anotherexample, data source 1802 can be connected to computing device 1850 by a cable, a directwireless link, and so on. Additionally or alternatively, in some configurations, data source1802 can be located locally and / or remotely from computing device 1850, and cancommunicate data to computing device 1850 (and / or server 1852) via a communicationnetwork (e.g., communication network 1854).

[0152] In some configurations, communication network 1854 can be any suitablecommunication network or combination of communication networks. For example,communication network 1854 can include a Wi-Fi network (which can include one ormore wireless routers, one or more switches, etc.), a peer-to-peer network (e.g., aBluetooth network), a cellular network (e.g., a 3G network, a 4G network, etc., complyingwith any suitable standard, such as CDMA, GSM, LTE, LTE Advanced, WiMAX, etc.), othertypes of wireless network, a wired network, and so on. In some configurations,communication network 1854 can be a local area network, a wide area network, a publicnetwork (e.g., the Internet), a private or semi-private network (e.g., a corporate oruniversity intranet), any other suitable type of network, or any suitable combination ofnetworks. Communications links shown in FIG. 18 can each be any suitablecommunications link or combination of communications links, such as wired links, fiberoptic links, Wi-Fi links, Bluetooth links, cellular links, and so on.32QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582

[0153] Referring now to FIG.19, an example of hardware 1900 that can be used toimplement data source 1802, computing device 1850, and server 1852 in accordancewith some configurations of the systems and methods described in the present disclosureis shown.

[0154] As shown in FIG. 19, in some configurations, computing device 1850 caninclude a processor 1902, a display 1904, one or more inputs 1906, one or morecommunication systems 1908, and / or memory 1910. In some configurations, processor1902 can be any suitable hardware processor or combination of processors, such as acentral processing unit (“CPU”), a graphics processing unit (“GPU”), and so on. In someconfigurations, display 1904 can include any suitable display devices, such as a liquidcrystal display (“LCD”) screen, a light-emitting diode (“LED”) display, an organic LED(“OLED”) display, an electrophoretic display (e.g., an “e-ink” display), a computermonitor, a touchscreen, a television, and so on. In some configurations, inputs 1906 caninclude any suitable input devices and / or sensors that can be used to receive user input,such as a keyboard, a mouse, a touchscreen, a microphone, and so on.

[0155] In some configurations, communications systems 1908 can include anysuitable hardware, firmware, and / or software for communicating information overcommunication network 1854 and / or any other suitable communication networks. Forexample, communications systems 1908 can include one or more transceivers, one ormore communication chips and / or chip sets, and so on. In a more particular example,communications systems 1908 can include hardware, firmware, and / or software thatcan be used to establish a Wi-Fi connection, a Bluetooth connection, a cellular connection,an Ethernet connection, and so on.

[0156] In some configurations, memory 1910 can include any suitable storagedevice or devices that can be used to store instructions, values, data, or the like, that canbe used, for example, by processor 1902 to present content using display 1904, tocommunicate with server 1852 via communications system(s) 1908, and so on. Memory1910 can include any suitable volatile memory, non-volatile memory, storage, or anysuitable combination thereof. For example, memory 1910 can include random-accessmemory (“RAM”), read-only memory (“ROM”), electrically programmable ROM(“EPROM”), electrically erasable ROM (“EEPROM”), other forms of volatile memory, otherforms of non-volatile memory, one or more forms of semi-volatile memory, one or moreflash drives, one or more hard disks, one or more solid state drives, one or more optical33QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582drives, and so on. In some configurations, memory 1910 can have encoded thereon, orotherwise stored therein, a computer program for controlling operation of computingdevice 1850. In such configurations, processor 1902 can execute at least a portion of thecomputer program to present content (e.g., images, user interfaces, graphics, tables),receive content from server 1852, transmit information to server 1852, and so on. Forexample, the processor 1902 and the memory 1910 can be configured to perform themethods described herein.

[0157] In some configurations, server 1852 can include a processor 1912, adisplay 1914, one or more inputs 1916, one or more communications systems 1918,and / or memory 1920. In some configurations, processor 1912 can be any suitablehardware processor or combination of processors, such as a CPU, a GPU, and so on. Insome configurations, display 1914 can include any suitable display devices, such as anLCD screen, LED display, OLED display, electrophoretic display, a computer monitor, atouchscreen, a television, and so on. In some configurations, inputs 1916 can include anysuitable input devices and / or sensors that can be used to receive user input, such as akeyboard, a mouse, a touchscreen, a microphone, and so on.

[0158] In some configurations, communications systems 1918 can include anysuitable hardware, firmware, and / or software for communicating information overcommunication network 1854 and / or any other suitable communication networks. Forexample, communications systems 1918 can include one or more transceivers, one ormore communication chips and / or chip sets, and so on. In a more particular example,communications systems 1918 can include hardware, firmware, and / or software thatcan be used to establish a Wi-Fi connection, a Bluetooth connection, a cellular connection,an Ethernet connection, and so on.

[0159] In some configurations, memory 1920 can include any suitable storagedevice or devices that can be used to store instructions, values, data, or the like, that canbe used, for example, by processor 1912 to present content using display 1914, tocommunicate with one or more computing devices 1850, and so on. Memory 1920 caninclude any suitable volatile memory, non-volatile memory, storage, or any suitablecombination thereof. For example, memory 1920 can include RAM, ROM, EPROM,EEPROM, other types of volatile memory, other types of non-volatile memory, one ormore types of semi-volatile memory, one or more flash drives, one or more hard disks,one or more solid state drives, one or more optical drives, and so on. In some34QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582configurations, memory 1920 can have encoded thereon a server program for controllingoperation of server 1852. In such configurations, processor 1912 can execute at least aportion of the server program to transmit information and / or content (e.g., data, images,a user interface) to one or more computing devices 1850, receive information and / orcontent from one or more computing devices 1850, receive instructions from one or moredevices (e.g., a personal computer, a laptop computer, a tablet computer, a smartphone),and so on.

[0160] In some configurations, the server 1852 is configured to perform themethods described in the present disclosure. For example, the processor 1912 andmemory 1920 can be configured to perform the methods described herein.

[0161] In some configurations, data source 1802 can include a processor 1922,one or more data acquisition systems 1924, one or more communications systems 1926,and / or memory 1928. In some configurations, processor 1922 can be any suitablehardware processor or combination of processors, such as a CPU, a GPU, and so on. Insome configurations, the one or more data acquisition systems 1924 are generallyconfigured to acquire data, images, or both, and can include an MRI system. Additionallyor alternatively, in some configurations, the one or more data acquisition systems 1924can include any suitable hardware, firmware, and / or software for coupling to and / orcontrolling operations of an MRI system. In some configurations, one or more portions ofthe data acquisition system(s) 1924 can be removable and / or replaceable.

[0162] Note that, although not shown, data source 1802 can include any suitableinputs and / or outputs. For example, data source 1802 can include input devices and / orsensors that can be used to receive user input, such as a keyboard, a mouse, atouchscreen, a microphone, a trackpad, a trackball, and so on. As another example, datasource 1802 can include any suitable display devices, such as an LCD screen, an LEDdisplay, an OLED display, an electrophoretic display, a computer monitor, a touchscreen,a television, etc., one or more speakers, and so on.

[0163] In some configurations, communications systems 1926 can include anysuitable hardware, firmware, and / or software for communicating information tocomputing device 1850 (and, in some configurations, over communication network 1854and / or any other suitable communication networks). For example, communicationssystems 1926 can include one or more transceivers, one or more communication chipsand / or chip sets, and so on. In a more particular example, communications systems 192635QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582can include hardware, firmware, and / or software that can be used to establish a wiredconnection using any suitable port and / or communication standard (e.g., VGA, DVI video,USB, RS-232, etc.), Wi-Fi connection, a Bluetooth connection, a cellular connection, anEthernet connection, and so on.

[0164] In some configurations, memory 1928 can include any suitable storagedevice or devices that can be used to store instructions, values, data, or the like, that canbe used, for example, by processor 1922 to control the one or more data acquisitionsystems 1924, and / or receive data from the one or more data acquisition systems 1924;to generate images from data; present content (e.g., data, images, a user interface) usinga display; communicate with one or more computing devices 1850; and so on. Memory1928 can include any suitable volatile memory, non-volatile memory, storage, or anysuitable combination thereof. For example, memory 1928 can include RAM, ROM,EPROM, EEPROM, other types of volatile memory, other types of non-volatile memory,one or more types of semi-volatile memory, one or more flash drives, one or more harddisks, one or more solid state drives, one or more optical drives, and so on. In someconfigurations, memory 1928 can have encoded thereon, or otherwise stored therein, aprogram for controlling operation of medical image data source 1802. In suchconfigurations, processor 1922 can execute at least a portion of the program to generateimages, transmit information and / or content (e.g., data, images, a user interface) to oneor more computing devices 1850, receive information and / or content from one or morecomputing devices 1850, receive instructions from one or more devices (e.g., a personalcomputer, a laptop computer, a tablet computer, a smartphone, etc.), and so on.

[0165] In some configurations, any suitable computer-readable media can be usedfor storing instructions for performing the functions and / or processes described herein.For example, in some configurations, computer-readable media can be transitory or non-transitory. For example, non-transitory computer-readable media can include mediasuch as magnetic media (e.g., hard disks, floppy disks), optical media (e.g., compact discs,digital video discs, Blu-ray discs), semiconductor media (e.g., RAM, flash memory,EPROM, EEPROM), any suitable media that is not fleeting or devoid of any semblance ofpermanence during transmission, and / or any suitable tangible media. As anotherexample, transitory computer-readable media can include signals on networks, in wires,conductors, optical fibers, circuits, or any suitable media that is fleeting and devoid of anysemblance of permanence during transmission, and / or any suitable intangible media.36QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582

[0166] As used herein in the context of computer implementation, unlessotherwise specified or limited, the terms “component,” “system,” “module,” “framework,”and the like are intended to encompass part or all of computer-related systems thatinclude hardware, software, a combination of hardware and software, or software inexecution. For example, a component may be, but is not limited to being, a processordevice, a process being executed (or executable) by a processor device, an object, anexecutable, a thread of execution, a computer program, or a computer. By way ofillustration, both an application running on a computer and the computer can be acomponent. One or more components (or system, module, and so on) may reside withina process or thread of execution, may be localized on one computer, may be distributedbetween two or more computers or other processor devices, or may be included withinanother component (or system, module, and so on).

[0167] In some implementations, devices or systems disclosed herein can beutilized or installed using methods embodying aspects of the disclosure.Correspondingly, description herein of particular features, capabilities, or intendedpurposes of a device or system is generally intended to inherently include disclosure of amethod of using such features for the intended purposes, a method of implementing suchcapabilities, and a method of installing disclosed (or otherwise known) components tosupport these purposes or capabilities. Similarly, unless otherwise indicated or limited,discussion herein of any method of manufacturing or using a particular device or system,including installing the device or system, is intended to inherently include disclosure, asembodiments of the disclosure, of the utilized features and implemented capabilities ofsuch device or system.

[0168] As used herein, the phrase "at least one of A, B, and C" means at least oneof A, at least one of B, and / or at least one of C, or any one of A, B, or C or combination ofA, B, or C. A, B, and C are elements of a list, and A, B, and C may be anything contained inthe present disclosure.

[0169] It is to be understood that the phraseology and terminology used herein isfor the purpose of description and should not be regarded as limiting. The use of“including,” “comprising,” or “having” and variations thereof herein is meant toencompass the items listed thereafter and equivalents thereof as well as additional items.Unless specified or limited otherwise, the terms “mounted,” “connected,” “supported,”and “coupled” and variations thereof are used broadly and encompass both direct and37QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.00582indirect mountings, connections, supports, and couplings. Further, “connected” and“coupled” are not restricted to physical or mechanical connections or couplings.

[0170] The present disclosure has described one or more preferred embodiments,and it should be appreciated that many equivalents, alternatives, variations, andmodifications, aside from those expressly stated, are possible and within the scope of theinvention.38QB\920171.00582\89858376.4

Claims

UMN 2022-135Quarles Ref.920171.00582CLAIMS1. A method for spatially encoding magnetic resonance signals duringmagnetic resonance imaging (MRI), the method comprising:generating, with an MRI system, a radio frequency (RF) excitation pulse to excitespins in a region-of-interest;generating, with the MRI system, a first frequency-modulated RF pulse todephase the spins excited by the RF excitation pulse, wherein the first frequency-modulated RF pulse is spatially varying in the region-of-interest;generating, with the MRI system, a second frequency-modulated RF pulse torephase the spins dephased by the first frequency-modulated RF pulse, wherein thesecond frequency-modulated RF pulse is spatially varying in the region-of-interest; andacquiring k-space data by sampling at least one echo signal generated whilerephasing the spins with the second frequency-modulated RF pulse.

2. The method of claim 1, wherein the first frequency-modulated RF pulsehas a first phase and the second frequency-modulated RF pulse has a second phase thatis different from the first phase.

3. The method of claim 2, wherein the second phase is 180 degrees differentthan the first phase.

4. The method of claim 1, wherein the first frequency-modulated RF pulsehas a first time-bandwidth product and the second frequency-modulated RF pulse has asecond time-bandwidth product that is different from the first time-bandwidth product.

5. The method of claim 4, wherein the second time-bandwidth product istwice the first time-bandwidth product.

6. The method of claim 1, wherein at least one of the first frequency-modulated RF pulse or the second frequency-modulated RF pulse comprises anadiabatic full passage (AFP) RF pulse.39QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.005827. The method of claim 1, wherein at least one of the first frequency-modulated RF pulse or the second frequency-modulated RF pulse comprise a hyperbolicsecant (HS) RF pulse.

8. The method of claim 1, wherein the second frequency-modulated RFpulse comprises a gapped transmission comprising at least two transmit periods duringwhich one or more respective segments of the second frequency-modulated RF pulseare generated and at least one gap period during which transmission of the secondfrequency-modulated RF pulse with the MRI system is turned off.

9. The method of claim 8, wherein the k-space data are acquired bysampling the at least one echo signal during the at least one gap period.

10. The method of claim 9, wherein the at least one echo signal comprises arotary echo.

11. The method of claim 1, further comprising reconstructing an image fromthe k-space data.

12. The method of claim 11, wherein reconstructing the image comprisesgenerating transformed k-space data by applying a rotation matrix to the acquired k-space data, wherein the image is reconstructed from the transformed k-space data.

13. The method of claim 12, wherein the rotation matrix is based on aneffective field angle to rotate the k-space data into a rotating reference frame, whereinthe effective field angle is defined based on a ratio between an amplitude modulationand a frequency modulation of the second frequency-modulated RF pulse.

14. The method of claim 1, wherein at least one of the first frequency-modulated RF pulse or the second frequency-modulated RF pulse is spatially varying inthe region-of-interest with a linear phase profile.40QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.0058215. The method of claim 1, wherein the k-space data are acquired bysampling the at least one echo signal generated while simultaneously transmitting thesecond frequency-modulated RF pulse.

16. The method of claim 15, wherein the k-space data are acquired usingsignal detection in a Rabi frequency range.

17. The method of claim 1, wherein the first frequency-modulated RF pulseand the second frequency-modulated RF pulse are spatially varying along a firstdimension, and wherein the method further comprises:generating, with the MRI system, a third frequency-modulated RF pulse toprovide a spatially varying phase of the spins along a second dimension;generating, with the MRI system, a fourth frequency-modulated RF pulse toprovide a spatially varying phase of the spins along the second dimension;18. The method of claim 17, wherein the first dimension is orthogonal to thesecond dimension.

19. A system for spatially encoding magnetic resonance signals duringmagnetic resonance imaging (MRI), the system comprising:a magnet system configured to produce a main magnetic field; anda radio frequency (RF) coil arrangement configured to:generate an RF excitation pulse to excite spins in a region-of -interest;generate a first frequency-modulated RF pulse to dephase the spinsexcited by the RF excitation pulse, wherein the first frequency-modulated RFpulse is spatially varying in the region-of-interest along a first dimension;generate a second frequency-modulated RF pulse to rephase the spinsexcited by the RF excitation pulse, wherein the second frequency-modulated RFpulse is spatially varying in the region-of-interest along the first dimension; andreceive at least one echo signal generated while rephasing the spins withthe second frequency-modulated RF pulse.41QB\920171.00582\89858376.4UMN 2022-135Quarles Ref.920171.0058220. The system of claim 19, further comprising a second RF coil arrangementconfigured to:generate a third frequency-modulated RF pulse to provide spatiallyvarying phase of the spins along a second dimension; andgenerate a fourth frequency-modulated RF pulse to provide spatiallyvarying phase of the spins along the second dimension.

21. The system of claim 20, wherein the first dimension is orthogonal to the seconddimension.^^ 42QB\920171.00582\89858376.4

Citation Information

Patent Citations

  • Ablation therapy using chemical shift magnetic resonance imaging

    US20040166062A1

  • Suppression of background tissues in MRA images

    US20080194944A1

  • System and method for producing radiofrequency pulses in magnetic resonance using an optimal phase surface

    US20200341084A1

  • Spin resonance signal measurement using a modulated fictitious field

    US20220397621A1

Cited By

  • Magnetic resonance imaging using two-dimensional phased-encoded frequency-modulated RABI encoded echoes

    WO2026043892A1

  • MRI using an off-resonant RF pulse to cause a phase shift or spoiling

    WO2026043893A1