DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 6/4/2026 has been entered.
Applicant' s arguments, 6/4/2026, have been fully considered. The following rejections and/or objections are either reiterated or newly applied. They constitute the complete set presently being applied to the instant application.
Claims 1-21 are the pending with claims 1-4, 6-8, 10-13, 15-17, and 19-21 being under examination. Claims 5, 9, 14, and 18 have previously been withdrawn.
Claim Interpretation
Claim 1 recites “each coil of the plurality of coils in the coil array being supplied with a different direct current of the first set of direct currents.” Under the broadest reasonable interpretation consistent with the specification, this limitation requires each coil of the recited plurality of coils to be supplied with a respective direct current having a current value different from the current value supplied to the other coils of that plurality. The specification explains that the DC currents are “determined for each coil” in the MC coil array, that “different coils in the shim array receive different DC currents 46, each having a predetermined magnitude,” and that “[t]he set of DC currents generally includes different values for each coil in the array” (Spec., ¶[0041], [0042], [0050]). Accordingly, identical current values applied to separately controlled coils do not constitute “different direct current[s]” merely because the coils or current channels are different. A difference in current value may include a difference in magnitude or polarity. The recited “plurality of coils,” however, need not include every coil of the coil array. Thus, the limitation requires different respective current values among the coils making up the recited plurality, but does not require every coil physically present in the coil array to receive a unique current value.
Claim 1 further recites that “the second set of direct currents [is] different than the first set of direct currents.” Under the broadest reasonable interpretation, the second set must differ from the first set with respect to at least one current value supplied to the recited plurality of coils. The claim does not require every corresponding current value in the second set to differ from the first set. A difference in magnitude or polarity of at least one corresponding current is sufficient. Merely characterizing identical current values as having a different spatial weighting or purpose would not make the current sets different. This interpretation is consistent with the specification's distinction between DC currents used during labeling to create a heterogeneous field and DC currents calculated and applied during imaging to improve field homogeneity (Spec., ¶[0041]-[0042], [0053]). The requirement that each coil of the recited plurality receive a respective different current value applies to the first set of direct currents. Claim 1 does not impose the same pairwise-different requirement on the second set.
Claim 1 applies the first set of direct currents to “a plurality of coils in the coil array” and subsequently applies the second set of direct currents to “the plurality of coils in the coil array.” Accordingly, the second set is applied to the same recited plurality of coils that receives the first set. The claim does not, however, require each coil of that plurality to contribute equally, or to contribute the same relative amount, to the resulting magnetic field during the first and second current sets. The respective field contribution of each coil may differ between the labeling and imaging operations as the current values supplied to the coils are changed. Thus, the same recited plurality may receive a first current vector optimized for the labeling plane and a second current vector optimized for the imaging plane without requiring each individual coil to provide an equally significant field contribution at both anatomical regions.
Claim 1 recites “a coil array configured for placement on a patient in the MRI system at a labeling plane.” Under the broadest reasonable interpretation consistent with the specification, this language requires the coil array to be configured for placement on the patient such that the array is capable of generating the recited magnetic field at the labeling plane. The claim does not require the individual coil conductors to lie physically within, or be geometrically coplanar with, the labeling plane. The specification describes a cervical ASL MC shim array that “can be placed in the labeling plane by placing on the patient's neck” and describes MC B0 shim arrays having loops positioned in close proximity to patient tissue to generate B0 offsets within the body (Spec., ¶[0029]-[0030]). The specification also describes placing a wearable shim coil array on or in contact with the patient's head or neck while controlling the field at the labeling plane (Spec., ¶[0048]).
Claim 1 recites a labeling plane that “bisects a plurality of arteries.” The term “plurality” requires two or more arteries and does not require the recited plurality to include every artery intersected by the labeling plane. Accordingly, two arteries intersected by the labeling plane may constitute the recited plurality even where additional arteries also intersect that plane.
Claim 1 further recites that the RF labeling field labels blood in at least one artery “while substantially not labeling” blood in the remainder of the recited plurality of arteries. Under the broadest reasonable interpretation consistent with the specification, “substantially not labeling” requires labeling efficiency sufficiently low that the blood does not meaningfully receive the intended spin labeling. The specification explains that blood in a region that is substantially off resonance “is said to not be substantially labeled by the RF field because the labeling efficiency is low,” and identifies labeling efficiency of less than 10 percent as the disclosed benchmark for such blood (Spec., ¶[0028]). The specification further identifies a region more than 100 Hz off resonance as one condition producing such low labeling efficiency. Claim 1 itself, however, does not expressly require the greater than 100 Hz condition, which is separately recited in dependent claim 6. Accordingly, the greater than 100 Hz condition is a disclosed manner of producing blood that is substantially not labeled, but is not imported as an additional numerical limitation into claim 1.
Claim 1 identifies the image sets as “first” and “second” MRI images and expressly requires the second MRI images to be captured “after a perfusion delay.” The claim does not state that the first and second MRI images must be obtained from the same immediately adjacent label and control pair, or that no intervening labeling or imaging operations may occur. Accordingly, to the extent the order of the recited steps is limiting, the first MRI images may be images acquired before the particular later labeling operation and second-image acquisition relied upon to satisfy the claim. This interpretation preserves the express temporal relationship associated with the perfusion delay without imposing an additional immediacy requirement not recited in the claim.
The foregoing interpretations of “each coil ... being supplied with a different direct current,” “the second set of direct currents being different than the first set of direct currents,” the antecedent relationship between “a plurality of coils” and “the plurality of coils,” “configured for placement ... at a labeling plane,” and “plurality of arteries” apply equally to corresponding language in claims 10 and 19. The interpretation of “substantially not labeling” applies to the corresponding limitation of claim 10. No other limitation of claims 1, 10, and 19 is thereby interpreted as coextensive.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 6, 10-13, 15-17, and 20 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 6 recites “nuclear spins” in line 2, but it is not clear if this recitation is the same as, related to, or different from “the nuclear spin” in claim 1, line 15. The relationship between these recitations should be clear.
Claim 10 recites “an MRI imaging system” in line 4, but it is not clear if this recitation is the same as, related to, or different from “A system for magnetic resonant imaging (MRI)” of claim 10, line 1. The relationship between these two recitations should be made clear.
Claims 11-13 and 15-17 are rejected by virtue of their dependence from claim 10.
Claim 15 recites “nuclear spins” in line 2, but it is not clear if this recitation is the same as, related to, or different from “the nuclear spin” in claim 10, lines 16-17. The relationship between these recitations should be clear.
Claim 20 recites “nuclear spins” in line 2, but it is not clear if this recitation is the same as, related to, or different from “a nuclear spin” in claim 19, lines 13-14. The relationship between these recitations should be clear.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Section 33(a) of the America Invents Act reads as follows:
Notwithstanding any other provision of law, no patent may issue on a claim directed to or encompassing a human organism.
Claim 13 is rejected under 35 U.S.C. 101 and section 33(a) of the America Invents Act as being directed to or encompassing a human organism. See also Animals - Patentability, 1077 Off. Gaz. Pat. Office 24 (April 21, 1987) (indicating that human organisms are excluded from the scope of patentable subject matter under 35 U.S.C. 101).
Claim 13 recites “wherein the coil array is a device placed on a head of the patient” in lines 1-2, which improperly include a human being as part of a claimed system.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-4, 6-8, 10-13, and 15-17 are rejected under 35 U.S.C. 103 as being unpatentable over van Harten et al. (T.W. van Harten et al., “On the ability to exploit signal fluctuations in pseudocontinuous arterial spin labeling for inferring the major flow territories from a traditional perfusion scan,” NeuroImage 230 (2021) 117813, available online January 29, 2021), hereinafter van Harten, and further in view of Jahanian et al. (Hesamoddin Jahanian et al., “B0 field inhomogeneity considerations in pseudo-continuous arterial spin labeling (pCASL): effects on tagging efficiency and correction strategy,” NMR in Biomedicine 24 (2011): 1202-1209), hereinafter Jahanian, and further in view of Stockmann et al. (Jason P. Stockmann et al., “Spatially-Selective Excitation Using a Tailored Nonlinear ΔB0 Pattern Generated by an Integrated Multi-Coil ΔB0/Rx Array,” Proceedings of the International Society for Magnetic Resonance in Medicine 26 (2018), Abstract 0170), hereinafter Stockmann (ISMRM), and further in view of Stockmann and Wald (Jason P. Stockmann and Lawrence L. Wald, “In Vivo B0 Field Shimming Methods for MRI at 7 T,” NeuroImage 168 (2018): 71-87), hereinafter Stockmann and Wald, and further in view of Cohen-Adad et al. (Julien Cohen-Adad et al., “32-Channel RF Coil Optimized for Brain and Cervical Spinal Cord at 3 T,” Magnetic Resonance in Medicine 66(4) (2011): 1198-1208), hereinafter Cohen-Adad, and further in view of Juchem et al. (Christoph Juchem et al., “Multi-Coil Shimming of the Mouse Brain,” Magnetic Resonance in Medicine 66(3) (2011): 893-900), hereinafter Juchem; as evidenced by Trampel et al. (Robert Trampel et al., “Continuous Arterial Spin Labeling Using a Local Magnetic Field Gradient Coil,” Magnetic Resonance in Medicine 48 (2002): 543-546), hereinafter Trampel.
Regarding claim 1, van Harten teaches a method for perfusion magnetic resonant imaging (MRI) within an MRI system having at least one RF coil configured to transmit an RF signal (van Harten, pp. 1-2: van Harten performs cerebral pCASL perfusion imaging on a 3 T MRI system using RF labeling pulses, wherein transmission of the disclosed RF labeling pulses necessarily employs an RF transmit coil of the MRI system).
Van Harten further teaches capturing one or more first MRI images of patient anatomy at a first imaging plane that is different from the labeling plane (van Harten, p. 2; Fig. 1: van Harten separately identifies the arterial labeling plane by time-of-flight angiography and acquires repeated pCASL control and labeled MRI data over a downstream brain imaging volume comprising 17 slices). Van Harten performs 35 repetitions, switches the position of optimal labeling to the contralateral internal carotid artery after each label-control pair, and splits the data according to which internal carotid artery was optimally labeled (van Harten, p. 2; Fig. 1 and accompanying caption at p. 3). Within a given artery-specific subset, control data from an earlier pair precede a later labeling operation and later labeled acquisition in that same subset. Those earlier control data constitute the first MRI images relied upon here.
Van Harten further teaches applying a series of RF pulses to the at least one RF coil, to create an RF labeling field in the labeling plane (van Harten, p. 2: van Harten applies pCASL labeling pulses and defines the “vessel-encoding distance” by the phase accumulated during the interpulse interval of the pCASL train, thereby establishing a labeling condition at the selected internal carotid artery).
Van Harten further teaches capturing one or more second MRI images of the patient anatomy at the first imaging plane after a perfusion delay (van Harten, p. 2: van Harten uses an 1800 ms labeling duration followed by an 1800 ms post-labeling delay and acquires the downstream 17-slice brain MRI data). The labeled data acquired in a later repetition of the same artery-specific subset, after the labeling period and post-labeling delay, constitute the second MRI images.
Van Harten further teaches comparing the one or more first MRI images and the one or more second MRI images to generate a map of perfusion associated with the at least one artery (van Harten, p. 2; Fig. 1 and accompanying caption at p. 3: van Harten states that optimal labeling is switched between the left and right internal carotid arteries after every label-control pair, that “the data were split into two,” and that the resulting mean ASL subtraction image is presented separately for the condition in which each internal carotid artery was optimally labeled; p. 4: van Harten further describes processing the repeated vessel-encoding image data by subtracting and averaging images according to the arterial labeling condition). Thus, van Harten expressly processes the repeated acquisitions as artery-specific image sets rather than limiting its perfusion result to a single isolated label-control pair. Within one such artery-specific set, the earlier control data identified above as the first MRI images precede the later labeling operation and later labeled data identified above as the second MRI images, and both are included in the repeated artery-specific data from which the corresponding mean ASL subtraction image is generated.
Also, regarding claim 1, van Harten does not fully teach providing a coil array configured for placement on a patient in the MRI system at a labeling plane that bisects a plurality of arteries; and applying a first set of direct currents to a plurality of coils in the coil array while the series of RF pulses are applied, each coil of the plurality of coils in the coil array being supplied with a different direct current of the first set of direct currents such that a first resulting field in the labeling plane is a non-uniform magnetic field such that the RF labeling field labels a nuclear spin of blood in at least one artery of the plurality of arteries, while substantially not labeling the nuclear spin of blood in a remainder of the plurality of arteries. Rather, van Harten teaches a labeling plane intersecting the left and right internal carotid arteries and intentionally establishes different labeling conditions at those arteries using an additional vessel-encoding gradient together with pCASL RF phase control. Van Harten states that the “vessel-encoding distance” is the distance producing a π phase shift during the interpulse interval and that, at that distance from the optimally labeled internal carotid artery, “the control and label condition will be switched compared to the targeted vessel” (van Harten, p. 2; Fig. 1), wherein the targeted ICA is placed in the label condition and the contralateral ICA is placed in the control condition. However, van Harten does not teach generating that artery-dependent spatial field with a patient-localized multi-coil B0 array in which the recited plurality of coils receive respective different DC current values during the RF labeling train.
Jahanian teaches that static B0 field inhomogeneity at the pCASL tagging plane changes the phase relationship between successive RF pulses and thereby changes labeling efficiency. Jahanian explains that field homogeneity in the tagging plane is often perturbed by the presence of the head or imperfect shimming and specifically teaches that air-tissue interfaces in the mouth and throat and dental work have “a profound effect” on the field at the tagging plane (Jahanian, p. 1203). Jahanian further teaches that the off-resonance ΔB0 term contributes to the position-dependent phase error between successive RF pulses and that the resulting phase error can degrade pCASL inversion (Jahanian, p. 1203, Eq. [2]). Jahanian teaches recalculating the RF phase for a selected static ΔB0 according to φlinear = γΔB0δ, thereby providing a known relationship for adjusting RF phase in view of a specified static B0 offset (Jahanian, p. 1203, Eq. [4]). Jahanian additionally states that negative tagging efficiencies represent situations in which “the order of tagging and control effectively changed” as a result of RF phase shifts greater than π (Jahanian, p. 1205; Fig. 4).
Jahanian further teaches that conventional scanner shimming may be inadequate at the pCASL tagging plane. Jahanian reports that standard and high-order shimming procedures applied over the whole brain, including the tagging and imaging planes, “were not successful” and in some instances higher-order shimming further distorted the field at the tagging or imaging planes (Jahanian, p. 1206). Jahanian generally seeks to correct unwanted B0 differences and teaches that an undesired difference in off-resonance between arteries may be eliminated. Jahanian is relied upon here for the disclosed physical relationship between static ΔB0, interpulse phase, RF phase adjustment, and pCASL labeling efficiency, and for its recognition that B0 control at the tagging plane may be inadequate using conventional shimming, rather than for Jahanian's preference as to whether a particular inter-artery difference should be created or removed. Van Harten independently supplies the reason to intentionally create the artery-dependent difference for vessel encoding.
Stockmann (ISMRM) teaches deliberately generating a nonlinear spatial ΔB0 pattern during RF excitation using independently driven MC ΔB0 loops. Stockmann (ISMRM) applies a min-max optimization that enforces frequency separation between target and excluded regions subject to maximum-current constraints of 3.5 A per coil or 40 A total, and reports that the resulting field “achieves spectral separation of the target and excluded regions” (Stockmann (ISMRM), pp. 1-2). Stockmann (ISMRM) further recognizes degraded performance where the coil loops do not sufficiently encircle the anatomy and states that this limitation could be overcome in future coil designs (Stockmann (ISMRM), p. 2). Stockmann (ISMRM) is relied upon for the MC field-synthesis mechanism during RF application and the demonstrated generation of spatially distinct resonance conditions, rather than for an assertion that its stationary-tissue selective excitation experiment itself teaches arterial spin labeling.
Stockmann and Wald teaches local MC B0 arrays comprising coils independently driven by DC currents to generate nonorthogonal field patterns and teaches solving a linear optimization problem for the currents that optimize the field within a selected ROI (Stockmann and Wald, p. 10). Stockmann and Wald further states that “by placing the coils relatively close to the head, B0 offsets in the brain large enough for 7 T shimming can be generated using modest current amplitudes on the order of a few amperes” (Stockmann and Wald, p. 10). Stockmann and Wald also expressly states that MC arrays “could be used for supplementary spatial encoding” and describes matrix-coil arrangements capable of generating flexible linear and nonlinear spatial encoding fields (Stockmann and Wald, p. 11).
Stockmann and Wald further teaches MC B0 field generation outside the brain. The review reports that “a 24ch MC shim array has recently been demonstrated for high-spatial order static and dynamic shimming of the spine” and explains that one approach to MC array design is to “target specific anatomy with a small number of coils” (Stockmann and Wald, p. 12).
Stockmann and Wald additionally teaches integrating B0 shim-field generation into an RF receive array when separate close-fitting MC and RF arrays would compete for space. The review explains that the integrated ΔB0/Rx approach unifies B0 shim production and RF receive functionality in the same physical coil winding by using inductive chokes to bridge DC current into the RF coil loop and across its RF tuning capacitors (Stockmann and Wald, p. 12). Stockmann and Wald further explains that a departure of the integrated ΔB0/Rx approach from previous MC shim arrays was to utilize the same close-fitting helmets used in RF arrays rather than cylindrical geometries (Stockmann and Wald, p. 13). The review identifies 8-channel and 31-channel B0 shim demonstration systems integrated into 3 T 32-channel RF receive arrays and reports prototype 3 T integrated ΔB0/Rx brain arrays based on conventional RF coil geometries, with only a modest RF SNR impact and no significant change in interelement noise correlation or parallel-imaging performance (Stockmann and Wald, pp. 12-13; Fig. 6c). The disclosed current-control hardware supplies up to 5 A per channel, accepts digital commands updating all shim settings in less than 1 ms, and can switch an individual low-inductance ΔB0/Rx element in less than 50 μs (Stockmann and Wald, p. 13).
Cohen-Adad teaches a close-fitting 32-channel RF loop array that “fully covers the brain and c-spine” (Cohen-Adad, p. 1, Abstract). Cohen-Adad further describes “a newly built close-fitting brain and c-spine coil designed to cover both regions with 32 circular loop elements,” wherein the design extends inferiorly to the T1-T2 level (Cohen-Adad, p. 2). The posterior former houses 30 array elements over the head and neck, with an optional two-element anterior neck paddle. The region covering the brain includes 20 overlapped circular loop elements having a 95 mm diameter, while the region covering the neck and cervical spine includes 10 overlapped circular loop elements having an 85 mm diameter (Cohen-Adad, p. 3; Fig. 1). Each element is a wire loop containing four distributed capacitors, including a tuning capacitor for operation at 3 T (Cohen-Adad, p. 3; Fig. 1d).
Cohen-Adad further demonstrates in vivo use of the array throughout the cervical region. Cohen-Adad performs diffusion imaging using eight slices covering C2-C7 and evaluates spinal-cord functional imaging over C2-C5 (Cohen-Adad, pp. 6-7). Cohen-Adad also demonstrates full brain and c-spine tractography using the same 32-channel array (Cohen-Adad, Fig. 7 and accompanying caption at p. 19). Thus, Cohen-Adad supplies a known patient-localized conventional RF loop geometry having physical loop coverage over both the cranial imaging region and the cervical region relevant to an arterial labeling plane.
Juchem teaches implementing MC field synthesis using individually controlled coil currents determined by constrained least-squares fitting. Juchem loads “a single set of 48 individual current values” into the MC interface and generates the corresponding currents using 48 individual amplifiers (Juchem, p. 4). Figure 1 identifies the resulting values as “coil-specific currents” and depicts at least a plurality of coils assigned mutually different current values (Juchem, Fig. 1 and accompanying caption at p. 12). Juchem further teaches current alterations over the full ±1 A range in as little as 10 μs, current stability better than ±50 ppm over an hour of continuous use, no significant coil-to-coil interactions or switching-related eddy currents, and negligible interaction between the scanner gradient system and the MC arrangement (Juchem, p. 4). Juchem additionally states that the individual currents used for static MC shimming “stayed on during the entire experiment” (Juchem, p. 4).
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified van Harten in view of Jahanian, Stockmann (ISMRM), Stockmann and Wald, Cohen-Adad, and Juchem by implementing van Harten's supplementary vessel-encoding function with the DC B0 field capability of an integrated ΔB0/Rx version of Cohen-Adad's patient-localized brain and cervical RF array, while retaining van Harten's ordinary pCASL labeling gradient, RF pulse train, and RF phase control. A person of ordinary skill would have implemented the modification by providing independently controlled DC paths through a fixed selected plurality of Cohen-Adad's conventional RF loop elements, including loop elements distributed in the cervical and cranial portions of the array, using the inductive-choke and current-control arrangement taught by Stockmann and Wald. Cohen-Adad's loops contain distributed tuning capacitors, and Stockmann and Wald expressly teaches bridging DC current into an RF loop and across its RF tuning capacitors, identifies integrated 3 T 32-channel RF receive arrays based on conventional RF geometries, and reports only a modest effect on RF performance (Stockmann and Wald, pp. 12-13; Cohen-Adad, p. 3).
A person of ordinary skill would have been motivated to provide that integrated construction because Stockmann and Wald expressly identifies competition for space between separate close-fitting MC and RF arrays and presents unification of B0 shim-field generation and RF reception in the same physical winding as a solution (Stockmann and Wald, p. 12). Applying that known integration technique to Cohen-Adad's existing close-fitting brain and cervical array would have avoided the need for a separate shim structure while providing independently controllable local B0 field generation over anatomy already covered by the RF array. This would have constituted application of Stockmann and Wald's known integrated ΔB0/Rx technique to the structurally similar conventional RF loop array of Cohen-Adad for the expressly identified benefit of combining RF reception and local B0 control without adding a competing close-fitting array.
A person of ordinary skill also would have had a specific reason to extend that local B0 control to the cervical tagging region. Jahanian identifies field perturbations near the mouth and throat as having a profound effect on the pCASL tagging plane and reports that conventional and high-order shimming over the tagging and imaging planes can be unsuccessful (Jahanian, pp. 1203, 1206). Stockmann and Wald, in turn, teaches that MC arrays may be designed to target specific anatomy and reports actual static and dynamic MC B0 shimming outside the brain using a 24-channel spine array (Stockmann and Wald, p. 12). These teachings would have provided a person of ordinary skill with a preexisting technical reason to provide independently controllable local B0 field generation in the cervical region covered by Cohen-Adad's array. Accordingly, a person of ordinary skill would have positioned the converted Cohen-Adad array on the patient with its cervical portion providing B0 field control at van Harten's arterial labeling plane intersecting the left and right internal carotid arteries, while the same physical array extended over the downstream cranial imaging region.
Having provided that independently controllable local B0 capability, a person of ordinary skill further would have had reason to use it for van Harten's supplementary vessel-encoding function. Van Harten expressly seeks to establish different labeling conditions at different feeding arteries so that their respective flow territories can be distinguished, while Stockmann and Wald expressly identifies MC arrays as useful for supplementary spatial encoding (van Harten, p. 2; Stockmann and Wald, p. 11). Using the MC field for that supplementary spatial-encoding function therefore would have applied an art-recognized spatial-encoding technique to accomplish van Harten's preexisting vessel-discrimination objective, rather than creating that objective from the claimed invention. Stockmann (ISMRM) further provides a known implementation directed to the same spatial-encoding objective by optimizing the MC currents to establish spectral separation between a selected target region and spatially separate excluded regions during RF excitation. A person of ordinary skill seeking to implement van Harten's vessel-discrimination objective using the MC spatial-encoding capability taught by Stockmann and Wald therefore would have had reason to use Stockmann (ISMRM)'s target-and-excluded-region optimization because it produces the spatially different RF resonance conditions required to distinguish the selected arterial region from the contralateral arterial region. A person of ordinary skill further would have had reason to use Juchem's constrained least-squares coil-current implementation because Juchem uses that procedure to determine the individual coil currents for synthesizing the desired MC B0 field.
A person of ordinary skill would have reasonably expected the modification to work because Jahanian establishes that static ΔB0 changes pCASL interpulse phase and labeling efficiency and demonstrates reversal of tag and control for sufficiently large phase shifts. This directly corresponds to van Harten's use of a π interpulse phase difference to place the targeted ICA in the label condition while the contralateral ICA is in the control condition. Jahanian's Eq. [4] further provides a known means for adjusting RF phase for the selected static ΔB0 condition, thereby preserving van Harten's pCASL RF phase control rather than requiring abandonment or reconstruction of it. Stockmann (ISMRM) demonstrates that independently controlled ΔB0/Rx loops can establish spectral separation between spatially distinct target and excluded regions during RF excitation, and Juchem provides a known implementation for synthesizing such fields with independently calculated coil-specific currents and maintaining a calculated current vector during MRI operation. Juchem's teaching that its individual currents remained on throughout the experiment further supports maintaining the first current set throughout van Harten's 1800 ms labeling train.
The required field scale likewise would not have rendered the modification technically unreasonable. Van Harten reports that 9.6 Hz of off-resonance produces a 4.14° phase shift during the 1.2 ms between its pCASL pulses (van Harten, p. 7). Because van Harten defines the contralateral vessel-encoding condition by a π, or 180°, interpulse phase difference, the corresponding frequency-offset scale is approximately 417 Hz, equivalent to a B0 difference of approximately 9.8 μT for protons. Stockmann and Wald teaches that close-fitting MC coils can generate B0 offsets sufficient for 7 T shimming using currents on the order of a few amperes and provides integrated current-control hardware capable of up to 5 A per channel, while Stockmann (ISMRM) actually achieves target-to-excluded spectral separation under finite per-coil and total-current constraints.
Trampel is additionally cited as evidence supporting the reasonable expectation that useful current-driven local magnetic-field generation at carotid anatomy was technically feasible at practical current levels. Trampel positions a local magnetic-field gradient module comprising current-driven loops at a human carotid labeling location and reports that the module generated a field gradient of approximately 0.5 mT/m/A at the expected location of the carotid artery (Trampel, pp. 543-544; Figs. 1-2). Trampel further reports that gradient strengths of 1.5 to 2 mT/m repeatedly produced greater than 90 percent inversion efficiency and that the corresponding currents in the gradient module were between approximately 3 and 4 A (Trampel, pp. 543-544). Trampel further reports that “the current in the gradient module remained switched on during the whole experiment, including shimming prior to the perfusion experiment” (Trampel, p. 545). This additionally evidences the practical feasibility of maintaining current-driven local magnetic-field generation at human carotid anatomy throughout an ASL experiment. Trampel's experimental arrangement also uses the local neck-mounted field-generating hardware together with separate cranial imaging hardware and expressly operates independently of the scanner gradient system (Trampel, pp. 543-545; Fig. 1). Trampel is relied upon as evidence that current-driven local coils positioned at carotid anatomy could practically generate substantial magnetic-field variation at practical current levels and remain energized throughout an ASL experiment. Trampel is not relied upon to establish the geometry or field direction of the proposed MC vessel-encoding field, and its measured gradient is not extrapolated across the spacing between the internal carotid arteries.
Thus, Stockmann and Wald and Stockmann (ISMRM) establish the known MC field-generation architecture and spatial field-control capability, while Trampel additionally evidences the practical feasibility of generating substantial current-driven magnetic-field variation at carotid anatomy. Using the left and right internal carotid arteries as the recited plurality of arteries, placing the cervical portion of the converted Cohen-Adad array so that its integrated loop elements provide field control at van Harten's arterial labeling plane, and selecting as the recited coil plurality a fixed plurality of independently driven loop elements assigned respective mutually different current values by the labeling optimization, a person of ordinary skill would have had a reasonable expectation that the first current set would generate a spatially nonuniform B0 field sufficient to maintain the selected internal carotid artery in the label condition while shifting the contralateral internal carotid artery to the corresponding control condition. Van Harten identifies that contralateral condition as control during the label acquisition. The resulting RF labeling operation would therefore label blood in the selected artery while substantially not labeling blood in the remainder of the selected arterial plurality.
Also, regarding claim 1, the modified van Harten does not fully teach that the one or more second MRI images previously mapped above are captured while applying a second set of direct currents to the plurality of coils in the coil array, the second set of direct currents being different than the first set of direct currents. Rather, the modified van Harten applies the first optimized current vector to the selected plurality of integrated loop elements during the labeling operation as discussed above. However, the modified van Harten does not yet teach applying a second current vector having at least one changed current value to that same selected plurality during acquisition of the second MRI images.
Stockmann and Wald teaches dynamic, imaging-region-specific MC shimming in which the MC current settings are changed during a multislice imaging experiment so that the B0 field can be optimized for the particular imaging slice rather than retaining a single global setting. Stockmann and Wald reports human 48-channel dynamic MC results in which the average σB0Global was reduced from 32.3 Hz using global first through third order shims to 13.3 Hz using dynamic MC shims, corresponding to a 59 percent improvement, and the EPI slices exhibited substantially reduced distortion. The MC shim settings were updated in under 1.5 ms during a 300 ms interslice delay without apparent artifacts (Stockmann and Wald, pp. 18-11; Fig. 8 and accompanying caption at pp. 32-33).
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have further modified the modified van Harten in view of Stockmann and Wald by retaining the same fixed selected plurality of integrated ΔB0/Rx loop elements used for the first current set, but changing the current vector applied to those same coil channels after completion of the artery-selective labeling operation so that a second imaging-specific current set having at least one changed current value is applied while the second MRI images are acquired.
The modification would have been technically feasible because Stockmann and Wald teaches calculating MC currents by optimization over a selected ROI and rapidly updating those current settings on the same MC hardware, while Cohen-Adad's single physical array extends from the cervical region through the brain. The fixed selected plurality includes loop elements distributed in the cervical and cranial portions of that integrated array. Accordingly, the same selected coil channels can be retained for both operations while their relative current values are changed according to the different spatial objectives. As interpreted above, the claim does not require every coil of the selected plurality to contribute equally, or to make the same relative contribution, to the magnetic field at both anatomical regions.
A person of ordinary skill would have been motivated to make this change because Stockmann and Wald expressly teaches that imaging-region-specific dynamic MC shimming improves B0 homogeneity and reduces EPI distortion in the slice being acquired (Stockmann and Wald, pp. 10-11; Fig. 8). Because the subsequent MRI data are acquired in a downstream imaging ROI different from the labeling region, a person of ordinary skill would have been motivated to reoptimize the already incorporated MC array for that imaging ROI to obtain Stockmann and Wald's expressly demonstrated improvements in B0 homogeneity and EPI distortion, rather than continue using a current vector optimized for the labeling region. The modification therefore would have constituted application of Stockmann and Wald's known dynamic MC shimming technique to the already incorporated MC hardware for its established imaging purpose. Van Harten provides an 1800 ms post-labeling delay, while Stockmann and Wald reports updating MC current settings in under 1.5 ms without apparent imaging artifacts. A person of ordinary skill therefore would have reasonably expected the second imaging-specific current set to be established before the subsequent image acquisition and maintained while the second MRI images are captured.
Regarding claim 2, the modified van Harten teaches the perfusion delay is 1-3 seconds after the applying the series of RF pulses (van Harten, p. 2: van Harten uses an 1800 ms labeling duration followed by an 1800 ms post-labeling delay before acquisition of the downstream 17-slice brain MRI data, wherein the 1800 ms post-labeling delay is 1.8 seconds and therefore falls within the recited range of 1-3 seconds).
Regarding claim 3, the modified van Harten teaches the capturing of the one or more first MRI images, the applying of the series of RF pulses, the applying of the first set of direct currents and the capturing of the one or more second MRI images are repeated for a second imaging plane (van Harten, p. 2: van Harten performs the pCASL acquisition over a 17-slice imaging volume with 35 repetitions; Fig. 1 and accompanying caption at p. 3: the position of optimal labeling is switched between the left and right internal carotid arteries after each label-control pair, confirming repeated label-control acquisitions throughout the scan). In the modified van Harten established regarding claim 1, each repeated labeling operation includes application of the first set of direct currents during the RF labeling train. Because the 35 repetitions repeatedly acquire corresponding pCASL image data at each of the same 17 imaging slices, a second one of those slices constitutes the recited second imaging plane. Across those repetitions, control data at that second imaging plane are repeatedly acquired, the RF labeling operation with the first set of direct currents is repeatedly performed, and labeled data at that second imaging plane are repeatedly acquired after the perfusion delay. Accordingly, the recited operations are repeated for the second imaging plane.
Regarding claim 4, the modified van Harten teaches the coil array is a device placed on a head of the patient (Cohen-Adad, pp. 1-3; Fig. 1: Cohen-Adad teaches a “close-fitting brain and c-spine coil” comprising a 32-channel loop array that fully covers the brain and cervical spinal cord, with the posterior former supporting array elements over the patient's head and neck; Stockmann and Wald, pp. 12-13: the integrated ΔB0/Rx approach incorporated regarding claim 1 utilizes the same close-fitting helmet geometries used for RF arrays). Thus, the integrated brain and cervical coil array incorporated into the modified van Harten is a device placed on the patient's head.
Regarding claim 6, the modified van Harten teaches the non-uniform magnetic field includes one or more regions where nuclear spins are at least 100 Hz off-resonance of the RF labeling field at a location of the remainder of the plurality of arteries (van Harten, p. 2: the contralateral internal carotid artery is placed in the control condition relative to the targeted internal carotid artery by a π interpulse phase difference; p. 7: 9.6 Hz of off-resonance produces a 4.14° phase shift during the disclosed 1.2 ms interpulse interval). As established regarding claim 1, the first current set implements van Harten's vessel-encoding condition using the MC field-synthesis mechanism of Stockmann (ISMRM), which establishes the target region on resonance with the RF excitation while spectrally separating the excluded region. Accordingly, the first current set maintains the selected internal carotid artery at the RF labeling resonance condition while providing van Harten's π interpulse phase separation at the contralateral internal carotid artery. Van Harten's disclosed relationship between off-resonance frequency and phase accumulation establishes that this π phase separation over the 1.2 ms interpulse interval corresponds to a frequency separation of approximately 417 Hz. The contralateral internal carotid artery, which constitutes the remainder of the selected plurality of arteries, is therefore approximately 417 Hz off-resonance of the RF labeling field and thus at least 100 Hz off-resonance.
Regarding claim 7, the modified van Harten teaches a second resulting field based on the second set of direct currents in the first imaging plane is a magnetic field that is more uniform than without the second set of direct currents (Stockmann and Wald, pp. 10-11; Fig. 8 and accompanying caption at pp. 32-33: dynamic imaging-region-specific MC shimming reduces the average σB0Global from 32.3 Hz using global first through third order shims without application of the dynamic MC current set to 13.3 Hz using dynamic MC shims, corresponding to a 59 percent improvement, and substantially reduces EPI distortion). As established regarding claim 1, the dynamic imaging-specific MC current vector constitutes the second set of direct currents applied during acquisition of the second MRI images. Accordingly, application of the second set produces a more uniform magnetic field in the first imaging plane than the field present without application of that second set.
Regarding claim 8, the modified van Harten teaches the at least one artery includes a carotid artery (van Harten, p. 2; Fig. 1: van Harten selectively labels the left and right internal carotid arteries and switches the position of optimal labeling between the two arteries after each label-control pair). As established regarding claim 1, the selected internal carotid artery constitutes the recited at least one artery.
Regarding claim 10, van Harten teaches a system for magnetic resonant imaging (MRI), the system comprising at least one RF coil configured to transmit an RF signal (van Harten, pp. 1-2: van Harten performs cerebral pCASL perfusion imaging on a 3 T MRI system using RF labeling pulses, wherein transmission of the disclosed RF labeling pulses necessarily employs an RF transmit coil of the MRI system). Van Harten was available online January 29, 2021.
Van Harten further teaches capturing one or more first MRI images of patient anatomy at a first imaging plane that is different from the labeling plane (van Harten, p. 2; Fig. 1: van Harten separately identifies the arterial labeling plane by time-of-flight angiography and acquires repeated pCASL control and labeled MRI data over a downstream brain imaging volume comprising 17 slices). Van Harten performs 35 repetitions, switches the position of optimal labeling to the contralateral internal carotid artery after each label-control pair, and splits the data according to which internal carotid artery was optimally labeled (p. 2; Fig. 1 and accompanying caption at p. 3). Within a given artery-specific subset, control data from an earlier pair precede a later labeling operation and later labeled acquisition in that same subset. Those earlier control data constitute the first MRI images relied upon here.
Van Harten further teaches applying a series of RF pulses to the at least one RF coil, to create an RF labeling field in the labeling plane (van Harten, p. 2: van Harten applies pCASL labeling pulses and defines the “vessel-encoding distance” by the phase accumulated during the interpulse interval of the pCASL train, thereby establishing a labeling condition at the selected internal carotid artery).
Van Harten further teaches capturing one or more second MRI images of the patient anatomy at the first imaging plane after a perfusion delay (van Harten, p. 2: van Harten uses an 1800 ms labeling duration followed by an 1800 ms post-labeling delay and acquires the downstream 17-slice brain MRI data). The labeled data acquired in a later repetition of the same artery-specific subset, after the labeling period and post-labeling delay, constitute the second MRI images.
Van Harten further teaches comparing the one or more first MRI images and the one or more second MRI images to generate a map of perfusion associated with the at least one artery (van Harten, p. 2; Fig. 1 and accompanying caption at p. 3: van Harten states that optimal labeling is switched between the left and right internal carotid arteries after every label-control pair, that “the data were split into two,” and that the resulting mean ASL subtraction image is presented separately for the condition in which each internal carotid artery was optimally labeled; p. 4: van Harten further describes processing the repeated vessel-encoding image data by subtracting and averaging images according to the arterial labeling condition). Thus, van Harten expressly processes the repeated acquisitions as artery-specific image sets rather than limiting its perfusion result to a single isolated label-control pair. Within one such artery-specific set, the earlier control data identified above as the first MRI images precede the later labeling operation and later labeled data identified above as the second MRI images, and both are included in the repeated artery-specific data from which the corresponding mean ASL subtraction image is generated.
Also, regarding claim 10, van Harten does not fully teach a coil array configured for placement on a patient in the MRI imaging system at a labeling plane that bisects a plurality of arteries; and a computer configured to apply a first set of direct currents to a plurality of coils in the coil array while the series of RF pulses are applied, each coil of the plurality of coils in the coil array being supplied with a different direct current of the first set of direct currents such that a first resulting field in the labeling plane is a non-uniform magnetic field such that the RF labeling field labels a nuclear spin of blood in at least one artery of the plurality of arteries, while substantially not labeling the nuclear spin of blood in a remainder of the plurality of arteries. Rather, van Harten teaches a labeling plane intersecting the left and right internal carotid arteries and intentionally establishes different labeling conditions at those arteries using an additional vessel-encoding gradient together with pCASL RF phase control. Van Harten states that the “vessel-encoding distance” is the distance producing a π phase shift during the interpulse interval and that, at that distance from the optimally labeled internal carotid artery, “the control and label condition will be switched compared to the targeted vessel” (p. 2; Fig. 1), wherein the targeted ICA is placed in the label condition and the contralateral ICA is placed in the control condition. However, van Harten does not teach generating that artery-dependent spatial field with a patient-localized multi-coil B0 array in which the recited plurality of coils receive respective different DC current values during the RF labeling train.
Jahanian teaches that static B0 field inhomogeneity at the pCASL tagging plane changes the phase relationship between successive RF pulses and thereby changes labeling efficiency. Jahanian explains that field homogeneity in the tagging plane is often perturbed by the presence of the head or imperfect shimming and specifically teaches that air-tissue interfaces in the mouth and throat and dental work have “a profound effect” on the field at the tagging plane (Jahanian, p. 1203). Jahanian further teaches that the off-resonance ΔB0 term contributes to the position-dependent phase error between successive RF pulses and that the resulting phase error can degrade pCASL inversion (p. 1203, Eq. [2]). Jahanian teaches recalculating the RF phase for a selected static ΔB0 according to φlinear = γΔB0δ, thereby providing a known relationship for adjusting RF phase in view of a specified static B0 offset (p. 1203, Eq. [4]). Jahanian additionally states that negative tagging efficiencies represent situations in which “the order of tagging and control effectively changed” as a result of RF phase shifts greater than π (p. 1205; Fig. 4).
Jahanian further teaches that conventional scanner shimming may be inadequate at the pCASL tagging plane. Jahanian reports that standard and high-order shimming procedures applied over the whole brain, including the tagging and imaging planes, “were not successful” and in some instances higher-order shimming further distorted the field at the tagging or imaging planes (p. 1206). Jahanian generally seeks to correct unwanted B0 differences and teaches that an undesired difference in off-resonance between arteries may be eliminated. Jahanian is relied upon here for the disclosed physical relationship between static ΔB0, interpulse phase, RF phase adjustment, and pCASL labeling efficiency, and for its recognition that B0 control at the tagging plane may be inadequate using conventional shimming, rather than for Jahanian's preference as to whether a particular inter-artery difference should be created or removed. Van Harten independently supplies the reason to intentionally create the artery-dependent difference for vessel encoding.
Stockmann (ISMRM) teaches deliberately generating a nonlinear spatial ΔB0 pattern during RF excitation using independently driven MC ΔB0 loops. Stockmann (ISMRM) applies a min-max optimization that enforces frequency separation between target and excluded regions subject to maximum-current constraints of 3.5 A per coil or 40 A total, and reports that the resulting field “achieves spectral separation of the target and excluded regions” (Stockmann (ISMRM), pp. 1-2). Stockmann (ISMRM) further recognizes degraded performance where the coil loops do not sufficiently encircle the anatomy and states that this limitation could be overcome in future coil designs (p. 2). Stockmann (ISMRM) is relied upon for the MC field-synthesis mechanism during RF application and the demonstrated generation of spatially distinct resonance conditions, rather than for an assertion that its stationary-tissue selective excitation experiment itself teaches arterial spin labeling.
Stockmann and Wald teaches local MC B0 arrays comprising coils independently driven by DC currents to generate nonorthogonal field patterns and teaches solving a linear optimization problem for the currents that optimize the field within a selected ROI (Stockmann and Wald, p. 10). Stockmann and Wald further states that “by placing the coils relatively close to the head, B0 offsets in the brain large enough for 7 T shimming can be generated using modest current amplitudes on the order of a few amperes” (p. 10). Stockmann and Wald also expressly states that MC arrays “could be used for supplementary spatial encoding” and describes matrix-coil arrangements capable of generating flexible linear and nonlinear spatial encoding fields (p. 11).
Stockmann and Wald further teaches MC B0 field generation outside the brain. The review reports that “a 24ch MC shim array has recently been demonstrated for high-spatial order static and dynamic shimming of the spine” and explains that one approach to MC array design is to “target specific anatomy with a small number of coils” (p. 12).
Stockmann and Wald additionally teaches integrating B0 shim-field generation into an RF receive array when separate close-fitting MC and RF arrays would compete for space. The review explains that the integrated ΔB0/Rx approach unifies B0 shim production and RF receive functionality in the same physical coil winding by using inductive chokes to bridge DC current into the RF coil loop and across its RF tuning capacitors (p. 12). Stockmann and Wald further explains that a departure of the integrated ΔB0/Rx approach from previous MC shim arrays was to utilize the same close-fitting helmets used in RF arrays rather than cylindrical geometries (p. 13). The review identifies 8-channel and 31-channel B0 shim demonstration systems integrated into 3 T 32-channel RF receive arrays and reports prototype 3 T integrated ΔB0/Rx brain arrays based on conventional RF coil geometries, with only a modest RF SNR impact and no significant change in interelement noise correlation or parallel-imaging performance (pp. 12-13; Fig. 6c). The disclosed current-control hardware supplies up to 5 A per channel, accepts digital commands updating all shim settings in less than 1 ms, and can switch an individual low-inductance ΔB0/Rx element in less than 50 μs (p. 13).
Cohen-Adad teaches a close-fitting 32-channel RF loop array that “fully covers the brain and c-spine” (Cohen-Adad, p. 1, Abstract). Cohen-Adad further describes “a newly built close-fitting brain and c-spine coil designed to cover both regions with 32 circular loop elements,” wherein the design extends inferiorly to the T1-T2 level (p. 2). The posterior former houses 30 array elements over the head and neck, with an optional two-element anterior neck paddle. The region covering the brain includes 20 overlapped circular loop elements having a 95 mm diameter, while the region covering the neck and cervical spine includes 10 overlapped circular loop elements having an 85 mm diameter (p. 3; Fig. 1). Each element is a wire loop containing four distributed capacitors, including a tuning capacitor for operation at 3 T (p. 3; Fig. 1d).
Cohen-Adad further demonstrates in vivo use of the array throughout the cervical region. Cohen-Adad performs diffusion imaging using eight slices covering C2-C7 and evaluates spinal-cord functional imaging over C2-C5 (pp. 6-7). Cohen-Adad also demonstrates full brain and c-spine tractography using the same 32-channel array (Fig. 7 and accompanying caption at p. 19). Thus, Cohen-Adad supplies a known patient-localized conventional RF loop geometry having physical loop coverage over both the cranial imaging region and the cervical region relevant to an arterial labeling plane.
Juchem teaches implementing MC field synthesis using individually controlled coil currents determined by constrained least-squares fitting. Juchem loads “a single set of 48 individual current values” into the MC interface and generates the corresponding currents using 48 individual amplifiers (Juchem, p. 4). Figure 1 identifies the resulting values as “coil-specific currents” and depicts at least a plurality of coils assigned mutually different current values (Fig. 1 and accompanying caption at p. 12). Juchem further teaches current alterations over the full ±1 A range in as little as 10 μs, current stability better than ±50 ppm over an hour of continuous use, no significant coil-to-coil interactions or switching-related eddy currents, and negligible interaction between the scanner gradient system and the MC arrangement (p. 4). Juchem additionally states that the individual currents used for static MC shimming “stayed on during the entire experiment” (p. 4). Juchem further teaches computer control of the MC current hardware, wherein C-based software running on the scanner console's Linux computer controls the MC shim interface via a serial port, and slice-specific sets of MC currents are applied under control of the pulse program via real-time TTL pulses from the scanner (pp. 3-4).
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified van Harten in view of Jahanian, Stockmann (ISMRM), Stockmann and Wald, Cohen-Adad, and Juchem by implementing van Harten's supplementary vessel-encoding function with the DC B0 field capability of an integrated ΔB0/Rx version of Cohen-Adad's patient-localized brain and cervical RF array, while retaining van Harten's ordinary pCASL labeling gradient, RF pulse train, and RF phase control. A person of ordinary skill would have implemented the modification by providing independently controlled DC paths through a fixed selected plurality of Cohen-Adad's conventional RF loop elements, including loop elements distributed in the cervical and cranial portions of the array, using the inductive-choke and current-control arrangement taught by Stockmann and Wald. Cohen-Adad's loops contain distributed tuning capacitors, and Stockmann and Wald expressly teaches bridging DC current into an RF loop and across its RF tuning capacitors, identifies integrated 3 T 32-channel RF receive arrays based on conventional RF geometries, and reports only a modest effect on RF performance (Stockmann and Wald, pp. 12-13; Cohen-Adad, p. 3). The computer would have been configured to command the calculated coil-specific current settings in coordination with the MRI pulse sequence using the scanner-console computer and pulse-program control taught by Juchem. A person of ordinary skill would have been motivated to use Juchem's scanner-console and pulse-program control because Juchem's pulse-program control provides a known technique for synchronizing application of MC current sets with the corresponding MRI sequence operations. Applying that control to the modified van Harten would permit the calculated first current set to be applied during the RF labeling interval for which the artery-selective field is required.
A person of ordinary skill would have been motivated to provide that integrated construction because Stockmann and Wald expressly identifies competition for space between separate close-fitting MC and RF arrays and presents unification of B0 shim-field generation and RF reception in the same physical winding as a solution (Stockmann and Wald, p. 12). Applying that known integration technique to Cohen-Adad's existing close-fitting brain and cervical array would have avoided the need for a separate shim structure while providing independently controllable local B0 field generation over anatomy already covered by the RF array. This would have constituted application of Stockmann and Wald's known integrated ΔB0/Rx technique to the structurally similar conventional RF loop array of Cohen-Adad for the expressly identified benefit of combining RF reception and local B0 control without adding a competing close-fitting array.
A person of ordinary skill also would have had a specific reason to extend that local B0 control to the cervical tagging region. Jahanian identifies field perturbations near the mouth and throat as having a profound effect on the pCASL tagging plane and reports that conventional and high-order shimming over the tagging and imaging planes can be unsuccessful (Jahanian, pp. 1203, 1206). Stockmann and Wald, in turn, teaches that MC arrays may be designed to target specific anatomy and reports actual static and dynamic MC B0 shimming outside the brain using a 24-channel spine array (Stockmann and Wald, p. 12). These teachings would have provided a person of ordinary skill with a preexisting technical reason to provide independently controllable local B0 field generation in the cervical region covered by Cohen-Adad's array. Accordingly, a person of ordinary skill would have positioned the converted Cohen-Adad array on the patient with its cervical portion providing B0 field control at van Harten's arterial labeling plane intersecting the left and right internal carotid arteries, while the same physical array extended over the downstream cranial imaging region.
Having provided that independently controllable local B0 capability, a person of ordinary skill further would have had reason to use it for van Harten's supplementary vessel-encoding function. Van Harten expressly seeks to establish different labeling conditions at different feeding arteries so that their respective flow territories can be distinguished, while Stockmann and Wald expressly identifies MC arrays as useful for supplementary spatial encoding (van Harten, p. 2; Stockmann and Wald, p. 11). Using the MC field for that supplementary spatial-encoding function therefore would have applied an art-recognized spatial-encoding technique to accomplish van Harten's preexisting vessel-discrimination objective, rather than creating that objective from the claimed invention. Stockmann (ISMRM) further provides a known implementation directed to the same spatial-encoding objective by optimizing the MC currents to establish spectral separation between a selected target region and spatially separate excluded regions during RF excitation. A person of ordinary skill seeking to implement van Harten's vessel-discrimination objective using the MC spatial-encoding capability taught by Stockmann and Wald therefore would have had reason to use Stockmann (ISMRM)'s target-and-excluded-region optimization because it produces the spatially different RF resonance conditions required to distinguish the selected arterial region from the contralateral arterial region. A person of ordinary skill further would have had reason to use Juchem's constrained least-squares coil-current implementation because Juchem uses that procedure to determine the individual coil currents for synthesizing the desired MC B0 field.
A person of ordinary skill would have reasonably expected the modification to work because Jahanian establishes that static ΔB0 changes pCASL interpulse phase and labeling efficiency and demonstrates reversal of tag and control for sufficiently large phase shifts. This directly corresponds to van Harten's use of a π interpulse phase difference to place the targeted ICA in the label condition while the contralateral ICA is in the control condition. Jahanian's Eq. [4] further provides a known means for adjusting RF phase for the selected static ΔB0 condition, thereby preserving van Harten's pCASL RF phase control rather than requiring abandonment or reconstruction of it. Stockmann (ISMRM) demonstrates that independently controlled ΔB0/Rx loops can establish spectral separation between spatially distinct target and excluded regions during RF excitation, and Juchem provides a known implementation for synthesizing such fields with independently calculated coil-specific currents and maintaining a calculated current vector during MRI operation. Juchem's teaching that its individual currents remained on throughout the experiment further supports maintaining the first current set throughout van Harten's 1800 ms labeling train.
The required field scale likewise would not have rendered the modification technically unreasonable. Van Harten reports that 9.6 Hz of off-resonance produces a 4.14° phase shift during the 1.2 ms between its pCASL pulses (van Harten, p. 7). Because van Harten defines the contralateral vessel-encoding condition by a π, or 180°, interpulse phase difference, the corresponding frequency-offset scale is approximately 417 Hz, equivalent to a B0 difference of approximately 9.8 μT for protons. Stockmann and Wald teaches that close-fitting MC coils can generate B0 offsets sufficient for 7 T shimming using currents on the order of a few amperes and provides integrated current-control hardware capable of up to 5 A per channel, while Stockmann (ISMRM) actually achieves target-to-excluded spectral separation under finite per-coil and total-current constraints.
Trampel is additionally cited as evidence supporting the reasonable expectation that useful current-driven local magnetic-field generation at carotid anatomy was technically feasible at practical current levels. Trampel positions a local magnetic-field gradient module comprising current-driven loops at a human carotid labeling location and reports that the module generated a field gradient of approximately 0.5 mT/m/A at the expected location of the carotid artery (Trampel, pp. 543-544; Figs. 1-2). Trampel further reports that gradient strengths of 1.5-2 mT/m repeatedly produced greater than 90 percent inversion efficiency and that the corresponding currents in the gradient module were between approximately 3 to 4 A (pp. 543-544). Trampel further reports that “the current in the gradient module remained switched on during the whole experiment, including shimming prior to the perfusion experiment” (p. 545). This additionally evidences the practical feasibility of maintaining current-driven local magnetic-field generation at human carotid anatomy throughout an ASL experiment. Trampel's experimental arrangement also uses the local neck-mounted field-generating hardware together with separate cranial imaging hardware and expressly operates independently of the scanner gradient system (pp. 543-545; Fig. 1). Trampel is relied upon as evidence that current-driven local coils positioned at carotid anatomy could practically generate substantial magnetic-field variation at practical current levels and remain energized throughout an ASL experiment. Trampel is not relied upon to establish the geometry or field direction of the proposed MC vessel-encoding field, and its measured gradient is not extrapolated across the spacing between the internal carotid arteries.
Thus, Stockmann and Wald and Stockmann (ISMRM) establish the known MC field-generation architecture and spatial field-control capability, while Trampel additionally evidences the practical feasibility of generating substantial current-driven magnetic-field variation at carotid anatomy. Using the left and right internal carotid arteries as the recited plurality of arteries, placing the cervical portion of the converted Cohen-Adad array so that its integrated loop elements provide field control at van Harten's arterial labeling plane, and selecting as the recited coil plurality a fixed plurality of independently driven loop elements assigned respective mutually different current values by the labeling optimization, a person of ordinary skill would have had a reasonable expectation that the first current set would generate a spatially nonuniform B0 field sufficient to maintain the selected internal carotid artery in the label condition while shifting the contralateral internal carotid artery to the corresponding control condition. Van Harten identifies that contralateral condition as control during the label acquisition. The resulting RF labeling operation would therefore label blood in the selected artery while substantially not labeling blood in the remainder of the selected arterial plurality. The modified system would include a computer configured to apply that first current set during the RF labeling train through the scanner-console and pulse-program control taught by Juchem.
Also, regarding claim 10, the modified van Harten does not fully teach that the computer is configured to capture the one or more second MRI images previously mapped above while applying a second set of direct currents to the plurality of coils in the coil array, the second set of direct currents being different than the first set of direct currents. Rather, the modified van Harten applies the first optimized current vector to the selected plurality of integrated loop elements during the labeling operation as discussed above. However, the modified van Harten does not yet teach applying a second current vector having at least one changed current value to that same selected plurality during acquisition of the second MRI images.
Stockmann and Wald teaches dynamic, imaging-region-specific MC shimming in which the MC current settings are changed during a multislice imaging experiment so that the B0 field can be optimized for the particular imaging slice rather than retaining a single global setting. Stockmann and Wald reports human 48-channel dynamic MC results in which the average σB0Global was reduced from 32.3 Hz using global first through third order shims to 13.3 Hz using dynamic MC shims, corresponding to a 59 percent improvement, and the EPI slices exhibited substantially reduced distortion. The MC shim settings were updated in under 1.5 ms during a 300 ms interslice delay without apparent artifacts (Stockmann and Wald, pp. 10-11; Fig. 8 and accompanying caption at pp. 32-33).
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have further modified the modified van Harten in view of Stockmann and Wald by configuring the computer to retain the same fixed selected plurality of integrated ΔB0/Rx loop elements used for the first current set, but change the current vector applied to those same coil channels after completion of the artery-selective labeling operation so that a second imaging-specific current set having at least one changed current value is applied while the second MRI images are acquired.
The modification would have been technically feasible because Stockmann and Wald teaches calculating MC currents by optimization over a selected ROI and rapidly updating those current settings on the same MC hardware, while Cohen-Adad's single physical array extends from the cervical region through the brain. The fixed selected plurality includes loop elements distributed in the cervical and cranial portions of that integrated array. Accordingly, the same selected coil channels can be retained for both operations while their relative current values are changed according to the different spatial objectives. As interpreted above, the claim does not require every coil of the selected plurality to contribute equally, or to make the same relative contribution, to the magnetic field at both anatomical regions. Juchem's scanner-console computer and pulse-program control of MC current sets further demonstrates that the change between the programmed current sets would have been implemented under computer control.
A person of ordinary skill would have been motivated to make this change because Stockmann and Wald expressly teaches that imaging-region-specific dynamic MC shimming improves B0 homogeneity and reduces EPI distortion in the slice being acquired (Stockmann and Wald, pp. 10-11; Fig. 8). Because the subsequent MRI data are acquired in a downstream imaging ROI different from the labeling region, a person of ordinary skill would have been motivated to reoptimize the already incorporated MC array for that imaging ROI to obtain Stockmann and Wald's expressly demonstrated improvements in B0 homogeneity and EPI distortion, rather than continue using a current vector optimized for the labeling region. The modification therefore would have constituted application of Stockmann and Wald's known dynamic MC shimming technique to the already incorporated MC hardware for its established imaging purpose.
Van Harten provides an 1800 ms post-labeling delay, while Stockmann and Wald reports updating MC current settings in under 1.5 ms without apparent imaging artifacts. A person of ordinary skill therefore would have reasonably expected the second imaging-specific current set to be established before the subsequent image acquisition and maintained while the second MRI images are captured.
Regarding claim 11, the modified van Harten teaches the perfusion delay is 1-3 seconds after the series of RF pulses are applied (van Harten, p. 2: van Harten uses an 1800 ms labeling duration followed by an 1800 ms post-labeling delay before acquisition of the downstream 17-slice brain MRI data, wherein the 1800 ms post-labeling delay is 1.8 seconds and therefore falls within the recited range of 1-3 seconds).
Regarding claim 12, the modified van Harten teaches the computer is further configured to repeat the capturing of the one or more first MRI images, the applying of the series of RF pulses, the applying of the first set of direct currents and the capturing of the one or more second MRI images for a second imaging plane (van Harten, p. 2: van Harten performs the pCASL acquisition over a 17-slice imaging volume with 35 repetitions; Fig. 1 and accompanying caption at p. 3: the position of optimal labeling is switched between the left and right internal carotid arteries after each label-control pair, confirming repeated label-control acquisitions throughout the scan). In the modified van Harten established regarding claim 10, the computer controls the repeated pCASL sequence, including the repeated RF labeling operations with application of the first set of direct currents and the corresponding image acquisitions. Because the 35 repetitions repeatedly acquire corresponding pCASL image data at each of the same 17 imaging slices, a second one of those slices constitutes the recited second imaging plane. Across those repetitions, the computer repeatedly acquires control data at that second imaging plane, performs the RF labeling operation with the first set of direct currents, and acquires labeled data at that second imaging plane after the perfusion delay. Accordingly, the computer is configured to repeat the recited operations for the second imaging plane.
Regarding claim 13, the modified van Harten teaches the coil array is a device placed on a head of the patient (Cohen-Adad, pp. 1-3; Fig. 1: Cohen-Adad teaches a “close-fitting brain and c-spine coil” comprising a 32-channel loop array that fully covers the brain and cervical spinal cord, with the posterior former supporting array elements over the patient's head and neck; Stockmann and Wald, pp. 12-13: the integrated ΔB0/Rx approach incorporated regarding claim 10 utilizes the same close-fitting helmet geometries used for RF arrays). Thus, the integrated brain and cervical coil array incorporated into the modified van Harten is a device placed on the patient's head.
Regarding claim 15, the modified van Harten teaches the non-uniform magnetic field includes one or more regions where nuclear spins are at least 100 Hz off-resonance of the RF labeling field at a location of the remainder of the plurality of arteries (van Harten, p. 2: the contralateral internal carotid artery is placed in the control condition relative to the targeted internal carotid artery by a π interpulse phase difference; p. 7: 9.6 Hz of off-resonance produces a 4.14° phase shift during the disclosed 1.2 ms interpulse interval). As established regarding claim 10, the first current set implements van Harten's vessel-encoding condition using the MC field-synthesis mechanism of Stockmann (ISMRM), which establishes the target region on resonance with the RF excitation while spectrally separating the excluded region. Accordingly, the first current set maintains the selected internal carotid artery at the RF labeling resonance condition while providing van Harten's π interpulse phase separation at the contralateral internal carotid artery. Van Harten's disclosed relationship between off-resonance frequency and phase accumulation establishes that this π phase separation over the 1.2 ms interpulse interval corresponds to a frequency separation of approximately 417 Hz. The contralateral internal carotid artery, which constitutes the remainder of the selected plurality of arteries, is therefore approximately 417 Hz off-resonance of the RF labeling field and thus at least 100 Hz off-resonance.
Regarding claim 16, the modified van Harten teaches a second resulting field based on the second set of direct currents in the first imaging plane is a magnetic field that is more uniform than without the second set of direct currents (Stockmann and Wald, pp. 10-11; Fig. 8 and accompanying caption at pp. 32-33: dynamic imaging-region-specific MC shimming reduces the average σB0Global from 32.3 Hz using global first through third order shims without application of the dynamic MC current set to 13.3 Hz using dynamic MC shims, corresponding to a 59 percent improvement, and substantially reduces EPI distortion). As established regarding claim 10, the dynamic imaging-specific MC current vector constitutes the second set of direct currents applied during acquisition of the second MRI images. Accordingly, application of the second set produces a more uniform magnetic field in the first imaging plane than the field present without application of that second set.
Regarding claim 17, the modified van Harten teaches the at least one artery includes a carotid artery (van Harten, p. 2; Fig. 1: van Harten selectively labels the left and right internal carotid arteries and switches the position of optimal labeling between the two arteries after each label-control pair). As established regarding claim 10, the selected internal carotid artery constitutes the recited at least one artery.
Claims 19-20 are rejected under 35 U.S.C. 103 as being unpatentable over van Harten et al. (T.W. van Harten et al., “On the ability to exploit signal fluctuations in pseudocontinuous arterial spin labeling for inferring the major flow territories from a traditional perfusion scan,” NeuroImage 230 (2021) 117813, available online January 29, 2021), hereinafter van Harten, and further in view of Jahanian et al. (Hesamoddin Jahanian et al., “B0 field inhomogeneity considerations in pseudo-continuous arterial spin labeling (pCASL): effects on tagging efficiency and correction strategy,” NMR in Biomedicine 24 (2011): 1202-1209), hereinafter Jahanian, and further in view of Stockmann and Wald (Jason P. Stockmann and Lawrence L. Wald, “In Vivo B0 Field Shimming Methods for MRI at 7 T,” NeuroImage 168 (2018): 71-87), hereinafter Stockmann and Wald, and further in view of Cohen-Adad et al. (Julien Cohen-Adad et al., “32-Channel RF Coil Optimized for Brain and Cervical Spinal Cord at 3 T,” Magnetic Resonance in Medicine 66(4) (2011): 1198-1208), hereinafter Cohen-Adad, and further in view of Juchem et al. (Christoph Juchem et al., “Multi-Coil Shimming of the Mouse Brain,” Magnetic Resonance in Medicine 66(3) (2011): 893-900), hereinafter Juchem; as evidenced by Trampel et al. (Robert Trampel et al., “Continuous Arterial Spin Labeling Using a Local Magnetic Field Gradient Coil,” Magnetic Resonance in Medicine 48 (2002): 543-546), hereinafter Trampel.
Regarding claim 19, van Harten teaches a method for perfusion magnetic resonant imaging (MRI) within an MRI system having at least one RF coil configured to transmit an RF signal (van Harten, pp. 1-2: van Harten performs cerebral pCASL perfusion imaging on a 3 T MRI system using RF labeling pulses, wherein transmission of the disclosed RF labeling pulses necessarily employs an RF transmit coil of the MRI system). Van Harten was available online January 29, 2021.
Van Harten further teaches capturing one or more first MRI images of patient anatomy at a first imaging plane that is different from the labeling plane (van Harten, p. 2; Fig. 1: van Harten separately identifies the arterial labeling plane by time-of-flight angiography and acquires repeated pCASL control and labeled MRI data over a downstream brain imaging volume comprising 17 slices). Van Harten performs 35 repetitions, switches the position of optimal labeling to the contralateral internal carotid artery after each label-control pair, and splits the data according to which internal carotid artery was optimally labeled (p. 2; Fig. 1 and accompanying caption at p. 3). Within a given artery-specific subset, control data from an earlier pair precede a later labeling operation and later labeled acquisition in that same subset. Those earlier control data constitute the first MRI images relied upon here.
Van Harten further teaches applying a series of RF pulses to the at least one RF coil, to create an RF labeling field in the labeling plane (van Harten, p. 2: van Harten applies pCASL labeling pulses and defines the “vessel-encoding distance” by the phase accumulated during the interpulse interval of the pCASL train, thereby establishing a labeling condition at the selected internal carotid artery).
Van Harten further teaches capturing one or more second MRI images of the patient anatomy at the first imaging plane after a perfusion delay (van Harten, p. 2: van Harten uses an 1800 ms labeling duration followed by an 1800 ms post-labeling delay and acquires the downstream 17-slice brain MRI data). The labeled data acquired in a later repetition of the same artery-specific subset, after the labeling period and post-labeling delay, constitute the second MRI images.
Van Harten further teaches comparing the one or more first MRI images and the one or more second MRI images to generate a map of perfusion (van Harten, p. 2; Fig. 1 and accompanying caption at p. 3: van Harten states that optimal labeling is switched between the left and right internal carotid arteries after every label-control pair, that “the data were split into two,” and that the resulting mean ASL subtraction image is presented separately for the condition in which each internal carotid artery was optimally labeled; p. 4: van Harten further describes processing the repeated vessel-encoding image data by subtracting and averaging images according to the arterial labeling condition). Thus, van Harten expressly processes the repeated acquisitions as artery-specific image sets rather than limiting its perfusion result to a single isolated label-control pair. Within one such artery-specific set, the earlier control data identified above as the first MRI images precede the later labeling operation and later labeled data identified above as the second MRI images, and both are included in the repeated artery-specific data from which the corresponding mean ASL subtraction image is generated.
Also, regarding claim 19, van Harten does not fully teach providing a coil array configured for placement on a patient in the MRI system at a labeling plane that bisects a plurality of arteries; and applying a first set of direct currents to a plurality of coils in the coil array while the series of RF pulses are applied, each coil of the plurality of coils in the coil array being supplied with a different direct current of the first set of direct currents such that a first resulting field in the labeling plane results in a labeling region where a nuclear spin of blood is more on-resonance with the RF labeling field than without the first set of direct currents. Rather, van Harten teaches an arterial labeling plane intersecting the left and right internal carotid arteries and performs pCASL labeling at that plane. However, van Harten does not teach using a patient-localized multi-coil B0 array in which the recited plurality of coils receive respective different DC current values during the RF labeling train to modify the static B0 field so that blood in a labeling region is more on-resonance with the RF labeling field than without those currents.
Jahanian teaches that static B0 field inhomogeneity at the pCASL tagging plane changes the phase relationship between successive RF pulses and thereby changes labeling efficiency. Jahanian explains that field homogeneity in the tagging plane is often perturbed by the presence of the head or imperfect shimming and specifically teaches that air-tissue interfaces in the mouth and throat and dental work have “a profound effect” on the field at the tagging plane (Jahanian, p. 1203). Jahanian further teaches that the off-resonance ΔB0 term contributes to the position-dependent phase error between successive RF pulses and that the resulting phase error can degrade pCASL inversion (p. 1203, Eq. [2]).
Jahanian further teaches that conventional scanner shimming may be inadequate at the pCASL tagging plane. Jahanian reports that standard and high-order shimming procedures applied over the whole brain, including the tagging and imaging planes, “were not successful” and in some instances higher-order shimming further distorted the field at the tagging or imaging planes (p. 1206). Thus, Jahanian identifies unwanted static B0 offset at the pCASL tagging plane as a known problem affecting the relationship between the local nuclear-spin resonance condition and the RF labeling field.
Stockmann and Wald teaches local MC B0 arrays comprising coils independently driven by DC currents to generate nonorthogonal field patterns and teaches solving a linear optimization problem for the currents that optimize the field within a selected ROI (Stockmann and Wald, p. 10). Stockmann and Wald further states that “by placing the coils relatively close to the head, B0 offsets in the brain large enough for 7 T shimming can be generated using modest current amplitudes on the order of a few amperes” (p. 10).
Stockmann and Wald further teaches MC B0 field generation outside the brain. The review reports that “a 24ch MC shim array has recently been demonstrated for high-spatial order static and dynamic shimming of the spine” and explains that one approach to MC array design is to “target specific anatomy with a small number of coils” (p. 12).
Stockmann and Wald additionally teaches integrating B0 shim-field generation into an RF receive array when separate close-fitting MC and RF arrays would compete for space. The review explains that the integrated ΔB0/Rx approach unifies B0 shim production and RF receive functionality in the same physical coil winding by using inductive chokes to bridge DC current into the RF coil loop and across its RF tuning capacitors (p. 12). Stockmann and Wald further explains that a departure of the integrated ΔB0/Rx approach from previous MC shim arrays was to utilize the same close-fitting helmets used in RF arrays rather than cylindrical geometries (p. 13). The review identifies 8-channel and 31-channel B0 shim demonstration systems integrated into 3 T 32-channel RF receive arrays and reports prototype 3 T integrated ΔB0/Rx brain arrays based on conventional RF coil geometries, with only a modest RF SNR impact and no significant change in interelement noise correlation or parallel-imaging performance (pp. 12-13; Fig. 6c). The disclosed current-control hardware supplies up to 5 A per channel, accepts digital commands updating all shim settings in less than 1 ms, and can switch an individual low-inductance ΔB0/Rx element in less than 50 μs (p. 13).
Cohen-Adad teaches a close-fitting 32-channel RF loop array that “fully covers the brain and c-spine” (Cohen-Adad, p. 1, Abstract). Cohen-Adad further describes “a newly built close-fitting brain and c-spine coil designed to cover both regions with 32 circular loop elements,” wherein the design extends inferiorly to the T1-T2 level (p. 2). The posterior former houses 30 array elements over the head and neck, with an optional two-element anterior neck paddle. The region covering the brain includes 20 overlapped circular loop elements having a 95 mm diameter, while the region covering the neck and cervical spine includes 10 overlapped circular loop elements having an 85 mm diameter (p. 3; Fig. 1). Each element is a wire loop containing four distributed capacitors, including a tuning capacitor for operation at 3 T (p. 3; Fig. 1d).
Cohen-Adad further demonstrates in vivo use of the array throughout the cervical region. Cohen-Adad performs diffusion imaging using eight slices covering C2-C7 and evaluates spinal-cord functional imaging over C2-C5 (pp. 6-7). Cohen-Adad also demonstrates full brain and c-spine tractography using the same 32-channel array (Fig. 7 and accompanying caption at p. 19). Thus, Cohen-Adad supplies a known patient-localized conventional RF loop geometry having physical loop coverage over both the cranial imaging region and the cervical region relevant to an arterial labeling plane.
Juchem teaches implementing MC field synthesis using individually controlled coil currents determined by constrained least-squares fitting. Juchem loads “a single set of 48 individual current values” into the MC interface and generates the corresponding currents using 48 individual amplifiers (Juchem, p. 4). Figure 1 identifies the resulting values as “coil-specific currents” and depicts at least a plurality of coils assigned mutually different current values (Fig. 1 and accompanying caption at p. 12). Juchem further teaches current alterations over the full ±1 A range in as little as 10 μs, current stability better than ±50 ppm over an hour of continuous use, no significant coil-to-coil interactions or switching-related eddy currents, and negligible interaction between the scanner gradient system and the MC arrangement (p. 4). Juchem additionally states that the individual currents used for static MC shimming “stayed on during the entire experiment” (p. 4).
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified van Harten in view of Jahanian, Stockmann and Wald, Cohen-Adad, and Juchem by providing the DC B0 field capability of an integrated ΔB0/Rx version of Cohen-Adad's patient-localized brain and cervical RF array and using independently controlled coil-specific currents to reduce static B0 offset in a labeling region at van Harten's arterial labeling plane while the pCASL RF labeling train is applied. A person of ordinary skill would have implemented the modification by providing independently controlled DC paths through a fixed selected plurality of Cohen-Adad's conventional RF loop elements, including loop elements distributed in the cervical and cranial portions of the array, using the inductive-choke and current-control arrangement taught by Stockmann and Wald. Cohen-Adad's loops contain distributed tuning capacitors, and Stockmann and Wald expressly teaches bridging DC current into an RF loop and across its RF tuning capacitors, identifies integrated 3 T 32-channel RF receive arrays based on conventional RF geometries, and reports only a modest effect on RF performance (Stockmann and Wald, pp. 12-13; Cohen-Adad, p. 3).
A person of ordinary skill would have been motivated to provide that integrated construction because Stockmann and Wald expressly identifies competition for space between separate close-fitting MC and RF arrays and presents unification of B0 shim-field generation and RF reception in the same physical winding as a solution (Stockmann and Wald, p. 12). Applying that known integration technique to Cohen-Adad's existing close-fitting brain and cervical array would have avoided the need for a separate shim structure while providing independently controllable local B0 field generation over anatomy already covered by the RF array. This would have constituted application of Stockmann and Wald's known integrated ΔB0/Rx technique to the structurally similar conventional RF loop array of Cohen-Adad for the expressly identified benefit of combining RF reception and local B0 control without adding a competing close-fitting array.
A person of ordinary skill also would have had a specific reason to extend that local B0 control to the cervical tagging region. Jahanian identifies field perturbations near the mouth and throat as having a profound effect on the pCASL tagging plane and reports that conventional and high-order shimming over the tagging and imaging planes can be unsuccessful (Jahanian, pp. 1203, 1206). Stockmann and Wald, in turn, teaches that MC arrays may be designed to target specific anatomy and reports actual static and dynamic MC B0 shimming outside the brain using a 24-channel spine array (Stockmann and Wald, p. 12). These teachings would have provided a person of ordinary skill with a preexisting technical reason to provide independently controllable local B0 field generation in the cervical region covered by Cohen-Adad's array. Accordingly, a person of ordinary skill would have positioned the converted Cohen-Adad array on the patient with its cervical portion providing B0 field control at van Harten's arterial labeling plane intersecting the left and right internal carotid arteries, while the same physical array extended over the downstream cranial imaging region.
Having provided that independently controllable local B0 capability, a person of ordinary skill further would have had reason to use it to reduce unwanted B0 offset in a labeling region at van Harten's arterial labeling plane. Jahanian expressly identifies tagging-plane B0 inhomogeneity as a source of interpulse phase error and degraded pCASL labeling and reports that conventional scanner shimming may inadequately correct the tagging-plane field (Jahanian, pp. 1203, 1206). Stockmann and Wald teaches optimizing independently driven MC currents for the B0 field within a selected ROI (Stockmann and Wald, p. 10). Applying that known MC shimming technique to a labeling region containing blood at van Harten's tagging plane would have reduced the local static B0 offset relative to the condition without the MC current set, thereby placing the nuclear spins of the blood more on-resonance with the RF labeling field. A person of ordinary skill seeking to perform that ROI-specific MC correction would have had reason to use Juchem's constrained least-squares coil-current implementation because Juchem determines individual coil currents for the purpose of synthesizing the desired MC B0 correction field, thereby providing a known implementation for producing the Stockmann and Wald ROI-optimized field with coil-specific current values.
A person of ordinary skill would have reasonably expected the modification to work because Jahanian establishes the relationship between static ΔB0 at the tagging plane, interpulse phase error, and pCASL labeling efficiency, while Stockmann and Wald establishes that independently driven MC arrays can generate B0 corrections optimized for a selected anatomical ROI. Juchem provides a known implementation for synthesizing such fields with independently calculated coil-specific currents and maintaining a calculated current vector during MRI operation. Juchem's teaching that its individual currents remained on throughout the experiment further supports maintaining the first current set throughout van Harten's 1800 ms labeling train.
Trampel is additionally cited as evidence supporting the reasonable expectation that useful current-driven local magnetic-field generation at carotid anatomy was technically feasible at practical current levels. Trampel positions a local magnetic-field gradient module comprising current-driven loops at a human carotid labeling location and reports that the module generated a field gradient of approximately 0.5 mT/m/A at the expected location of the carotid artery (Trampel, pp. 543-544; Figs. 1-2). Trampel further reports that gradient strengths of 1.5-2 mT/m repeatedly produced greater than 90 percent inversion efficiency and that the corresponding currents in the gradient module were between approximately 3 to 4 A (pp. 543-544). Trampel further reports that “the current in the gradient module remained switched on during the whole experiment, including shimming prior to the perfusion experiment” (p. 545). This additionally evidences the practical feasibility of maintaining current-driven local magnetic-field generation at human carotid anatomy throughout an ASL experiment. Trampel's experimental arrangement also uses the local neck-mounted field-generating hardware together with separate cranial imaging hardware and expressly operates independently of the scanner gradient system (pp. 543-545; Fig. 1). Trampel is relied upon as evidence that current-driven local coils positioned at carotid anatomy could practically generate substantial magnetic-field variation at practical current levels and remain energized throughout an ASL experiment.
Thus, placing the cervical portion of the converted Cohen-Adad array so that its integrated loop elements provide B0 field control at van Harten's arterial labeling plane and selecting as the recited coil plurality a fixed plurality of independently driven loop elements assigned respective mutually different current values by the labeling-region optimization, a person of ordinary skill would have had a reasonable expectation that the first current set would reduce unwanted static B0 offset in the labeling region. The resulting first field would therefore provide a labeling region in which the nuclear spins of blood are more on-resonance with the RF labeling field than without application of the first set of direct currents.
Also, regarding claim 19, the modified van Harten does not fully teach capturing one or more second MRI images of the patient anatomy at the first imaging plane after a perfusion delay while applying a second set of direct currents to the plurality of coils in the coil array, the second set of direct currents being different than the first set of direct currents. Rather, the modified van Harten teaches the post-labeling acquisition of the second MRI images as discussed above and applies the first MC current set during the labeling operation to improve the resonance condition in the labeling region. However, the modified van Harten does not teach changing at least one current value from the first set and applying the resulting second current set to the same selected plurality during acquisition of the second MRI images.
Stockmann and Wald teaches dynamic, imaging-region-specific MC shimming in which the MC current settings are changed during a multislice imaging experiment so that the B0 field is optimized for the particular imaging slice rather than retaining a single global setting. Stockmann and Wald reports human 48-channel dynamic MC results in which the average σB0Global was reduced from 32.3 Hz using global first through third order shims to 13.3 Hz using dynamic MC shims, corresponding to a 59 percent improvement, and the EPI slices exhibited substantially reduced distortion. The MC shim settings were updated in under 1.5 ms during a 300 ms interslice delay without apparent artifacts (Stockmann and Wald, pp. 10-11; Fig. 8 and accompanying caption at pp. 32-33).
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have further modified the modified van Harten in view of Stockmann and Wald by retaining the same fixed selected plurality of integrated ΔB0/Rx loop elements used for the first current set, but changing at least one current value after completion of the labeling operation so that a second imaging-specific MC current set is applied while the second MRI images are acquired.
The modification would have been technically feasible because Stockmann and Wald teaches calculating MC currents by optimization over a selected ROI and rapidly updating those current settings on the same MC hardware, while Cohen-Adad's single physical array extends from the cervical region through the brain. The fixed selected plurality includes loop elements distributed in the cervical and cranial portions of that integrated array. Accordingly, the same selected coil channels can be retained for both operations while their respective current values are changed according to the different spatial optimization regions. As interpreted above, the claim does not require every coil of the selected plurality to contribute equally, or to make the same relative contribution, to the magnetic field at both anatomical regions.
A person of ordinary skill would have been motivated to make this change because Stockmann and Wald expressly teaches that imaging-region-specific dynamic MC shimming improves B0 homogeneity and reduces EPI distortion in the slice being acquired (Stockmann and Wald, pp. 10-11; Fig. 8). The first current set is optimized for the B0 condition in the arterial labeling region, whereas the subsequent MRI data are acquired at a different, downstream imaging plane. A person of ordinary skill therefore would have been motivated to reoptimize the already incorporated MC array for the imaging ROI by changing at least one of the coil-current values rather than retaining a current vector optimized for a different anatomical region. The modification therefore would have constituted application of Stockmann and Wald's known dynamic MC shimming technique to the already incorporated MC hardware for its established imaging purpose.
Van Harten provides an 1800 ms post-labeling delay, while Stockmann and Wald reports updating MC current settings in under 1.5 ms without apparent imaging artifacts. A person of ordinary skill therefore would have reasonably expected the second imaging-specific current set to be established before the subsequent image acquisition and maintained while the second MRI images are captured.
Regarding claim 20, the modified van Harten does not fully teach the first resulting field in the labeling plane includes one or more regions where nuclear spins are at least 100 Hz off-resonance of the RF labeling field. Rather, as established regarding claim 19, the first set of direct currents is selected to improve the resonance condition of blood in a labeling region relative to the condition without the first set of direct currents. However, the modified van Harten does not yet teach configuring that first resulting field such that another region of the labeling plane is at least 100 Hz off-resonance of the RF labeling field.
Van Harten teaches that the contralateral internal carotid artery is placed in the control condition relative to the targeted internal carotid artery by a π interpulse phase difference (van Harten, p. 2; Fig. 1). Van Harten further teaches that 9.6 Hz of off-resonance produces a 4.14° phase shift during the disclosed 1.2 ms interpulse interval (p. 7). Accordingly, implementing van Harten's π, or 180°, contralateral vessel-encoding condition using a static B0 field requires a frequency separation of approximately 417 Hz between the resonance conditions at the targeted and contralateral internal carotid arteries.
Jahanian teaches that static ΔB0 at the pCASL tagging plane changes the phase relationship between successive RF pulses and thereby changes labeling efficiency, and provides the known relationship between static B0 offset and the resulting interpulse phase shift (Jahanian, p. 1203, Eqs. [2], [4]). Jahanian additionally teaches that sufficiently large phase shifts can reverse the tagging and control conditions (p. 1205; Fig. 4).
Stockmann and Wald teaches that MC arrays generate flexible spatial B0 fields from independently controlled DC currents and expressly states that MC arrays “could be used for supplementary spatial encoding,” including generation of flexible linear and nonlinear spatial encoding fields (Stockmann and Wald, pp. 10-11).
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have further modified the modified van Harten by calculating the first set of direct currents such that, while improving the resonance condition in the targeted labeling region as established regarding claim 19, the resulting spatial B0 field also provides van Harten's π vessel-encoding condition at the contralateral internal carotid artery. A person of ordinary skill would have been motivated to make the modification because van Harten expressly seeks to establish different labeling conditions at different feeding arteries so that their respective flow territories can be distinguished, while Stockmann and Wald expressly identifies MC arrays as useful for supplementary spatial encoding. Using the already incorporated MC field for that supplementary spatial-encoding function therefore would have applied an art-recognized spatial-encoding technique to accomplish van Harten's preexisting vessel-discrimination objective.
A person of ordinary skill would have reasonably expected the modification to work because Jahanian establishes that static ΔB0 changes pCASL interpulse phase and labeling efficiency and demonstrates reversal of tag and control for sufficiently large phase shifts. This directly corresponds to van Harten's use of a π interpulse phase difference to place the targeted internal carotid artery in the label condition while the contralateral internal carotid artery is in the control condition. As established regarding claim 19, the first MC current set is used to improve the resonance condition in the targeted labeling region. A person of ordinary skill seeking the improved labeling efficiency identified by Jahanian would have had reason to optimize that region to, or sufficiently near, resonance with the RF labeling field. Providing van Harten's approximately 417 Hz vessel-encoding frequency separation relative to that labeling region would place the contralateral region more than 100 Hz off-resonance of the RF labeling field. Accordingly, the first resulting field would include one or more regions where nuclear spins are at least 100 Hz off-resonance of the RF labeling field.
Claim 21 is rejected under 35 U.S.C. 103 as being unpatentable over van Harten et al. (T.W. van Harten et al., “On the ability to exploit signal fluctuations in pseudocontinuous arterial spin labeling for inferring the major flow territories from a traditional perfusion scan,” NeuroImage 230 (2021) 117813, available online January 29, 2021), hereinafter van Harten, and further in view of Jahanian et al. (Hesamoddin Jahanian et al., “B0 field inhomogeneity considerations in pseudo-continuous arterial spin labeling (pCASL): effects on tagging efficiency and correction strategy,” NMR in Biomedicine 24 (2011): 1202-1209), hereinafter Jahanian, and further in view of Stockmann et al. (Jason P. Stockmann et al., “Spatially-Selective Excitation Using a Tailored Nonlinear ΔB0 Pattern Generated by an Integrated Multi-Coil ΔB0/Rx Array,” Proceedings of the International Society for Magnetic Resonance in Medicine 26 (2018), Abstract 0170), hereinafter Stockmann (ISMRM), and further in view of Stockmann and Wald (Jason P. Stockmann and Lawrence L. Wald, “In Vivo B0 Field Shimming Methods for MRI at 7 T,” NeuroImage 168 (2018): 71-87), hereinafter Stockmann and Wald, and further in view of Cohen-Adad et al. (Julien Cohen-Adad et al., “32-Channel RF Coil Optimized for Brain and Cervical Spinal Cord at 3 T,” Magnetic Resonance in Medicine 66(4) (2011): 1198-1208), hereinafter Cohen-Adad, and further in view of Juchem et al. (Christoph Juchem et al., “Multi-Coil Shimming of the Mouse Brain,” Magnetic Resonance in Medicine 66(3) (2011): 893-900), hereinafter Juchem, as evidenced by Trampel et al. (Robert Trampel et al., “Continuous Arterial Spin Labeling Using a Local Magnetic Field Gradient Coil,” Magnetic Resonance in Medicine 48 (2002): 543-546), hereinafter Trampel, and further in view of Zhao et al. (Li Zhao et al., “Improving the Robustness of Pseudo Continuous Arterial Spin Labeling to Off Resonance and Pulsatile Flow Velocity,” Magnetic Resonance in Medicine 78(4) (2017): 1342-1351), hereinafter Zhao; .
The modified van Harten teaches the limitations of claim 1 as discussed above.
Regarding claim 21, the modified van Harten does not fully teach pausing the first set of direct currents while a main body coil of the MRI system applies background field suppression pulses. Van Harten performs its acquisition using a “standard background suppressed pCASL” protocol, thereby establishing that background suppression is already included in the pCASL acquisition relied upon for claim 1 (van Harten, p. 2). As established regarding claim 1, however, the first set of direct currents is applied during the pCASL RF labeling train to generate the artery-selective spatial B0 field. Stockmann (ISMRM) further teaches pulse-sequence control of the MC field, wherein the coil currents can be switched quickly and energized differently during different portions of the MRI sequence. Stockmann (ISMRM) controls the MC hardware from the MRI pulse sequence and turns the MC field off during a different sequence operation (Stockmann (ISMRM), p. 1; Fig. 2). Thus, the modified van Harten teaches background suppression and teaches selectively applying and discontinuing the MC field according to the MRI sequence operation, but does not teach pausing the first set of direct currents while a main body coil applies the background-suppression pulses.
Zhao teaches performing pCASL on a 3 T MRI scanner “using the body coil for transmission” and further teaches that “[i]nterleaved labeling and background suppression was used to reduce motion artifacts” (Zhao, p. 6, “Volunteer Experiments and Evaluations”). Because Zhao performs the pCASL acquisition using the body coil for RF transmission, the interleaved background-suppression RF pulses of that acquisition would likewise have been transmitted by the body coil. Zhao therefore provides the recited main-body-coil implementation of the background-suppression operation.
It would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have further modified the modified van Harten in view of Zhao by using the pulse-sequence-controlled MC switching taught by Stockmann (ISMRM) to pause the first set of direct currents while the body coil applies Zhao's interleaved background-suppression pulses and to restore the first set for the pCASL labeling intervals. A person of ordinary skill would have been motivated to make the modification to preserve the effectiveness of the background-suppression RF operation. As established regarding claim 1, the first current set is specifically calculated to create a spatially varying B0 field that establishes different resonance conditions at different arterial locations for vessel-selective labeling. Stockmann and Wald expressly teaches that ΔB0 can reduce the effectiveness of suppression and further teaches that ΔB0 has undesired effects on RF excitation pulses, including RF inversion pulses (Stockmann and Wald, p. 2). Stockmann (ISMRM) further demonstrates the mechanism by which the first current set deliberately establishes spatially different resonance conditions, calculating the MC field so that a given RF bandwidth is on-resonance within the target region but not outside that region (Stockmann (ISMRM), p. 1). Zhao, in contrast, interleaves background suppression with the labeling operation to reduce motion artifacts (Zhao, p. 6). A person of ordinary skill therefore would have had reason not to impose the labeling-specific spatial ΔB0 pattern during the separate background-suppression RF operation, because the art expressly recognized that ΔB0 can adversely affect suppression and RF inversion. Using the pulse-sequence-controlled switching already taught by Stockmann (ISMRM), the person of ordinary skill would instead have paused the first current set while the body coil applies the background-suppression pulses and restored the first current set for the labeling intervals, thereby retaining the spatially selective B0 pattern when needed for vessel-selective labeling without unnecessarily imposing that pattern during the separate suppression operation.
A person of ordinary skill would have reasonably expected the modification to be successfully implemented because Stockmann (ISMRM) expressly controls the MC hardware from the MRI pulse sequence and switches the MC current field between sequence operations. Stockmann and Wald further teaches that digital commands for updating all shim settings can be issued in less than 1 ms and that an individual low-inductance ΔB0/Rx element can be switched in less than 50 μs, while Juchem teaches alteration of individual coil currents over the full ±1 A range in as little as 10 μs (Stockmann and Wald, p. 13; Juchem, p. 4). Thus, the incorporated MC hardware was capable of pausing the first set during the background-suppression intervals and restoring the first set for the labeling operation within the timing of the MRI pulse sequence.
Response to Arguments
Objections
Applicant's arguments filed 6/4/2026, page 8, regarding the previous Objections to claim 3 have been fully considered and are persuasive. Applicant's amendment to claim 3 overcomes the previously identified informality. Accordingly, the previous objection to claim 3 is withdrawn.
Claim Interpretation
Upon reconsideration of the claim language in view of the Specification, the previous interpretations concerning “each coil ... being supplied with a different direct current” and “the second set of direct currents being different than the first set of direct currents” are superseded by the interpretations set forth above. In particular, the Office no longer relies on the previous interpretation that the first-current-set limitation is satisfied merely because at least some coil currents differ from at least some other coil currents, or that identical corresponding current values may constitute different first and second current sets based solely on different spatial weighting, purpose, or subset characterization.
35 U.S.C. §103
Applicant's arguments filed 6/4/2026, pages 8-11, regarding the previous rejections of claims 1-4, 6-8, 10-13, 15-17, and 19-20 under 35 U.S.C. 103 have been fully considered. The previous rejections of claims 1-4, 6-8, 10-13, 15-17, and 19-21 are not maintained. The present action sets forth new grounds of rejection based on materially different reference combinations, allocations of teachings, and obviousness rationales. Accordingly, Applicant's arguments directed to the previous grounds are moot to the extent they challenge those former grounds.
Applicant's Argument: Applicant argues that the previous combination of Stockmann, Zhao, and Juchem did not teach that application of the different direct currents to the plurality of coils produces the claimed non-uniform magnetic field such that the RF labeling field labels blood in at least one artery while substantially not labeling blood in the remainder of the recited plurality. Applicant argues that Helle does not remedy that deficiency because Helle obtains spatially selective labeling by dynamically changing transverse gradients in combination with changes to the RF pulse phase, rather than by the claimed different direct currents producing the artery-selective field. Applicant further notes that Hernandez-Garcia and Jahanian were not relied upon in the previous action to remedy that asserted deficiency. Applicant relies on the same reasons with respect to claims 2-4 and 6-8 by virtue of their dependency from claim 1, applies the same reasoning to claims 10 and 19, and relies on the corresponding dependency of claims 11-13, 15-17, and 20.
Examiner's Response: These arguments are moot with respect to the present rejections because the previous grounds are not maintained. The present action does not rely on Helle to establish artery-selective labeling resulting from the first set of direct currents and does not rely on the former allocation of teachings among Stockmann, Zhao, Juchem, Helle, Hernandez-Garcia, and Jahanian. Rather, the present action applies new grounds including van Harten as the primary reference and materially different supporting teachings and obviousness rationales, as set forth in detail in the rejections above. Helle and Hernandez-Garcia are not relied upon in the present rejections. To the extent Stockmann and Wald, Juchem, or Jahanian are relied upon in the present action, those references are applied as part of the newly stated grounds and for the teachings and rationales expressly identified above. Applicant's arguments concerning whether Helle cured the deficiency of the former combination, and whether that former combination connected Juchem's coil-specific currents to the claimed resulting field, therefore no longer require resolution with respect to the superseded grounds.
Applicant's arguments that claims 2-4, 6-8, 11-13, 15-17, and 20 are patentable by virtue of their dependency from claims 1, 10, or 19 are likewise moot as directed to the previous grounds. The present rejections of those claims are based on the new grounds set forth above.
Applicant identifies claim 21 among the claims rejected in the previous action but does not present a separate substantive argument concerning claim 21 or its additional limitation. The previous rejection of claim 21 is not maintained, and the present rejection of claim 21 is set forth above on a new ground.
Conclusion
Any inquiry concerning this communication or earlier communications from the
examiner should be directed to AARON MERRIAM whose telephone number is (703) 756-
5938. The examiner can normally be reached M-F 8:00 am - 5:00 pm.
Examiner interviews are available via telephone, in-person, and video conferencing
using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is
encouraged to use the USPTO Automated Interview Request (AIR) at
http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s
supervisor, Jason Sims can be reached on (571)272-4867. The fax phone number for the
organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained
from Patent Center. Unpublished application information in Patent Center is available to
registered users. To file and manage patent submissions in Patent Center, visit:
https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for
more information about Patent Center and https://www.uspto.gov/patents/docx for
information about filing in DOCX format. For additional questions, contact the Electronic
Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO
Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/AARON MERRIAM/Examiner, Art Unit 3791
/MATTHEW KREMER/Primary Examiner, Art Unit 3791
Prosecution Insights