WO2018050777A1 - Procédé de caractérisation du champ d'émission de fréquences radioélectriques dans une nmr - Google Patents

Procédé de caractérisation du champ d'émission de fréquences radioélectriques dans une nmr Download PDF

Info

Publication number
WO2018050777A1
WO2018050777A1 PCT/EP2017/073195 EP2017073195W WO2018050777A1 WO 2018050777 A1 WO2018050777 A1 WO 2018050777A1 EP 2017073195 W EP2017073195 W EP 2017073195W WO 2018050777 A1 WO2018050777 A1 WO 2018050777A1
Authority
WO
WIPO (PCT)
Prior art keywords
signals
acquired
signal
sample
sequence
Prior art date
Application number
PCT/EP2017/073195
Other languages
English (en)
Inventor
Mélina BOULDI
Jan WARNKING
Original Assignee
Institut National De La Sante Et De La Recherche Medicale (Inserm)
Universite Grenoble Alpes
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Institut National De La Sante Et De La Recherche Medicale (Inserm), Universite Grenoble Alpes filed Critical Institut National De La Sante Et De La Recherche Medicale (Inserm)
Publication of WO2018050777A1 publication Critical patent/WO2018050777A1/fr

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R33/00—Arrangements or instruments for measuring magnetic variables
    • G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
    • G01R33/24—Arrangements or instruments for measuring magnetic variables involving magnetic resonance for measuring direction or magnitude of magnetic fields or magnetic flux
    • G01R33/246—Spatial mapping of the RF magnetic field B1

Definitions

  • the invention relates to a method of measuring the radio-frequency magnetic field (Bi-field) transmitted by a nuclear magnetic resonance (NMR) apparatus.
  • Bi-field radio-frequency magnetic field
  • NMR nuclear magnetic resonance
  • MR magnetic resonance
  • MRI magnetic resonance imaging
  • a sample to be examined being for example a patient's body
  • Bo field uniform magnetic field
  • RF field radio frequency
  • Bi field of defined frequency
  • the transversal component of the nuclear magnetization generated by such an RF excitation precesses around the Bo field at a characteristic frequency called the Larmor frequency, and gives rise to a measureable RF magnetic field emitted by the sample which is at the origin of the magnetic resonance signal providing information about the sample.
  • a MR signal is usually acquired in response to an imaging sequence that performs imaging by encoding spatial information in the signal in a way to permit creating an at least two-dimensional map, i.e., image of the signal.
  • at least two MR signals can be acquired in response to an imaging sequence, wherein each signal contains complete spatial information to reconstruct an image or is represented by a reconstructed image.
  • the one or more MR signals are reconstructed to form a 2D or 3D MR image, where the intensity and possibly the phase of each voxel in the MR image represents the MR signal originating at least partially from a volume element at a corresponding imaged location in the sample, thus creating an at least two-dimensional map of the distribution of said signal in at least a portion of the sample.
  • the way in which properties of the sample, properties of the MRI apparatus and the interaction between the two affect the MR signal, and thus the contrast in the MR image depends on the MRI sequence used. MRI sequences can produce either a single MR signal of a certain contrast, represented by a single MR image, or several MR signals of identical or differing contrasts, which are usually reconstructed as separate MR images.
  • Bi mapping a reliable measurement of the spatial distribution of the transmitted RF field
  • Bi mapping usually requires the acquisition, in response to one or more imaging sequence, of at least two distinct MR signals, that is, two signals whose dependence on any one or several factors among sample-dependent factors, transmit Bi field strength and Bo field strength is different.
  • MR signals depend on many sample- and machine- dependent factors besides Bi, requiring analysis methods that render the resulting Bi map independent of at least a factor common to all acquired signals.
  • Bi can be derived from the signals.
  • some analysis methods explicitly normalize the MR signals, for example by calculating the ratio of MR signals. Analysis methods based on the phase of the acquired signals may be based on phase differences, eliminating a common phase factor. Still other analysis methods achieve independence of a common factor in a way that is implicit in the method.
  • a signal may have a cyclic i.e. periodic dependence on Bi, such as a signal acquired in response to an AFI sequence or a DREAM sequence.
  • Some Bi-mapping methods generate a signal that does not exhibit a periodic dependency on the local Bi, such as the Bloch-Siegert shift method as published by Sacolick et al.
  • This acquisition sequence leads to an MR signal whose phase is proportional to the square of the local Bi value. While the dynamic range is still limited, since phase values larger than 2 ⁇ are not distinguishable from smaller ones, the Bi increment that increases phase from 0 to 2 ⁇ is different from the Bi increment that increases phase from 2 ⁇ to 4 ⁇ and similarly for the subsequent phase intervals.
  • the acquired signals have non-cyclic dependences on BI and differ by the sign in the signal phase.
  • the method of Bi mapping with the shortest acquisition time is the DREAM method as disclosed in the above mentioned publication "dual refocusing echo acquisition mode " and application EP 2 615 470.
  • This method comprises emitting an imaging sequence including at least two preparation RF pulses and acquiring one or more free induction decay (FID) signals and one or more stimulated echo (SE) signals, from which a Bi map indicating the spatial distribution of the RF field is reconstructed.
  • FID free induction decay
  • SE stimulated echo
  • This method is limited to a measurable dynamic range of an actual flip angle of the preparation RF pulses inferior to about ⁇ 80° for obtaining an appropriate precision.
  • this sequence is capable of correctly representing a real maximal Bi about 33% higher than the nominal Bi .
  • a flip angle of about 30° should be used for the preparation RF pulses, which will reduce the SNR of the stimulated echo signal by a factor of about 3 and increase the uncertainty in the calculated Bi by about 70%.
  • the acquisition would need to be repeated three times, increasing acquisition time by the same amount.
  • a limited dynamic range may not be prohibitive in most applications involving homogeneous samples and weak static main field Bo because the range of the values of Bi met in these cases are relatively narrow, often varying by less than 50% compared to the nominal value.
  • the variations of transmit Bi field can be more important, for example in the presence of parallel-transmit systems where the transmit field produced by individual coils needs to be measured.
  • Systems capable of emitting the RF field using a plurality of radio frequency transmit coils and integrating adjustment programs to calibrate their respective RF transmit fields have been commercialized by several MRI vendors.
  • the Bi fields of these individual coils can be highly heterogeneous, sometimes varying by an order of magnitude or more between regions in the sample that are proximal to the individual element and regions that are distal.
  • the dynamic range required for Bi measurements can thus be large, unlike that for homogeneous volume-transmit coils.
  • Bi fields also become increasingly heterogeneous at higher magnetic field strengths, leading to a need for Bi mapping sequences with relatively large dynamic range.
  • Bi can be highly heterogeneous
  • an electrically conducting implant for example a deep brain stimulation (DBS) or pacemaker lead in a patient.
  • DBS deep brain stimulation
  • Any RF current induced in the implant by the externally applied RF field leads to a local Bi hot-spot close to the implant.
  • the Bi map close to the implant being directly related to induced RF current, disposing of a method to accurately map a highly heterogeneous RF transmit fields with a sequence of low RF power might permit to assess the safety of an MRI procedure at the beginning of an imaging session using such an appropriate Bi mapping pre-scan. This requires limiting the number of acquisitions needed to cover a given (large) dynamic range.
  • the flip angle of the RF pulses used can be reduced, increasing the period of the signal dependence on the local Bi relative to the nominal Bi and thus the range of Bi values that can be encoded in the signal, at the cost of lower signal intensity and thus lower precision in regions of low Bi .
  • Another solution to acquire Bi mapping data with large dynamic range is based on the acquisition of a plurality of Bi maps, with imaging sequences of different acquisition parameters and thus different dynamic ranges and precisions, followed by local use of signals acquired with the most appropriate sequences.
  • Bi mapping Another difficulty of Bi mapping is the dependence of the acquired MR signals on factors other than Bi, such as the relaxation time Ti of the sample, which may give rise to local variations of the signals which are not related to Bi .
  • Most Bi mapping methods are based on sequences tailored to provide signals or ratios of signals that are approximately independent of Ti and other confounding effects, followed by inversion of the approximate signal equation to reconstruct Bi, neglecting the above mentioned dependence.
  • This approach thus may impose other constraints on the acquisition parameters in order to obtain a reliable reconstruction of the Bi map from an approximate signal equation that may be valid only in a given regime (e.g. TR»Ti for classic dual-angle or multiple-angle methods; TR «Ti for the AFI sequence; low flip angle for relative Bi mapping using SPGR).
  • a separately acquired Ti map is integrated in the process of reconstruction of the Bi map. This approach is particularly useful if a Ti map is required in the given context, but it may not be optimal in terms of acquisition time when only Bi information is sought, since the acquisition time spent on the Ti map adds relatively little information about Bi .
  • Other Bi mapping methods were designed specifically to minimize such effects in the acquisition step, with other disadvantages such as very complex sequences, long acquisition times and/or high transmitted RF power.
  • MRF magnetic resonance imaging
  • the signal evolution acquired in a given sequence block depends on the acquisition parameters of previous sequence blocks.
  • the nuclear magnetizations of spin ensembles of given properties are not in a steady state during the imaging sequence, imparting a strong pseudo-random evolution on them which is specific to their relaxation and resonance properties and which can be described using numerical simulations of the Bloch equations.
  • Bi is among the parameters that can be modeled and extracted from MR fingerprinting data according to the article by Buonincontri et al, Magnetic Resonance in Medicine, in press, 2015, although the aim of such methods is mainly in avoiding bias in the estimation of other parameters by specifically modeling the impact of Bi variations on the signal. Since the aim and strength of MR fingerprinting is the measurement of multi-parametric data, since the signal is explicitly made to depend strongly on many different sample properties, there has not been a demonstration that MR fingerprinting sequences can be competitive with respect to dedicated Bi mapping sequences in dynamic range and/or SNR efficiency when specifically aiming only for Bi mapping.
  • WO 2009/1 18702 discloses a magnetic resonance method including performing an actual flip angle mapping (AFI) sequence, wherein all of the signals of the first and second imaging sequences exhibit a periodic dependency on Bl , all with identical periods.
  • AFI flip angle mapping
  • Magnetic Resonance in Medicine 2006, 56: 1375-1379 disclose methods that rely on a mixed SE / STE sequence which produces both a spin-echo (SE) and a stimulated echo (STE) signal.
  • SE spin-echo
  • STE stimulated echo
  • these methods repeat the acquisition a number of times using various flip angles.
  • the signals are periodic in Bi and the proposed flip angle variation produces a variation in the periodicity of the dependence of the signal on Bi.
  • the drawback of the method as used in Antoine Lutti et al. and Jiru et al. is the fact that they reconstruct individual Bl maps from the data acquired with each of the sequences, or at least, combine the two signals from each sequence prior to further combining data between sequences.
  • Such combination includes calculating a ratio between the acquired signals, which risks making the resulting Bl map or combined signal unstable if the signal in the denominator (the spin-echo signal) is small.
  • This either limits the method to a dynamic range corresponding to a ⁇ 180°, as in Jiru et al, or requires exclusion of some data from the processing, as in Lutti et al.
  • data that is excluded in this way, if not combined by calculating the ratio of the acquired signals is sensitive to Bi and may contribute to improving the precision of the Bi map if taken into account appropriately.
  • Magnetic Resonance in Medicine 77:229-2378 discloses a method to acquire Bl distribution plots that estimate the histogram of Bl values in a volume, by reconstructing a projection of Bl along a single spatial dimension, as an alternative to spatial Bl mapping methods.
  • the acquisition sequence allows localizing the signal in a single dimension only.
  • the present invention aims to further improve the method of Bi mapping, in particular to increase the dynamic range of Bi mapping, by providing a method for
  • N 2 is an integer equal or superior to 2
  • the first and second imaging sequences being chosen so that at least one signal of the first and second subsets of MR signals does not exhibit a cyclic dependency on Bi, or at least one signal of the first subset of MR signals and at least one signal of the second subset of MR signals have cyclic dependencies on Bi with periods that are different and not an integer multiple of each other,
  • acquiring a subset of magnetic resonance (MR) signals having a number equal or superior to 2, preferably distinct as having different dependence on Bl allows to maximize the amount of information about Bi that can be encoded in MR signals within a given acquisition time and to efficiently distinguish the effects of Bi from the effects of other factors on the signal in order to render the Bi map independent of at least a factor common to all acquired signals.
  • MR magnetic resonance
  • Disposing of at least one signal for which the dependences on Bi is not periodic or at least one pair of signals comprising a signal of the first subset of MR signals and a signal of the second subset of MR signals, and for which the dependences on Bi are periodic with periods that are not an integer multiple of each other is necessary to reconstruct a Bi map with a dynamic range exceeding that admitted by the signal of the longer periodicity, for example each of the individual sequences.
  • signals of the first subset are acquired before the acquisition of part or all the signals of the second subset.
  • Two signals having different dependences on Bi may be two signals that are periodic with different periods, two signals that are non-periodic with different dependence on Bi, one signal that is periodic and one that is not, or even two signals that are periodic with the same period but have different dependence on BI within that period, such as signals acquired in response to a standard AFI sequence.
  • the method may comprise subjecting the portion of the sample to one or more imaging sequences comprising RF pulses to acquire one or more subsets of magnetic resonance (MR) signals ([Si", S 2 ", SNS”], . . ..[SI 11 - 1 , S "1 , S 1"1 ]), after step b) and before step c) and wherein the map of the spatial distribution of the RF transmit field strength Bi produced by the nuclear magnetic resonance apparatus within the sample is derived on the basis of some or all of the MR signals and preferably, on the basis of a non-linear model of these signals as a function of at least Bi.
  • MR magnetic resonance
  • the acquired signals refer to MR signals acquired in response to the first and second imaging sequences in steps a) and b), i.e. the first and second subsets of MR signals, and if appropriate, the signals acquired in response to the one or more subsets of MR signals after step b) and before step c).
  • At least one signal of the one or more subsets of MR signals acquired after step b) and before step c) does not exhibit a cyclic dependency on Bi or exhibits a cyclic dependency on Bi with a period that is not an integer multiple of that of at least one signal of both the first and second subsets of MR signals.
  • Bi may be derived directly from the acquired signals at an imaged location, preferably at each imaged location, without linear or non-linear transformation of the acquired signals.
  • all the acquired signals of the first and second subset of signals may be simultaneously processed to derive the map of B i .
  • simultaneous processing of the signals it should be understood that part of the acquired signals of the first and second subset are not transformed for example via a linear or non-linear transformation as described above, independently of the others.
  • step c) comprises applying a linear or non-linear transformation to the acquired signal sets and deriving the map of the spatial distribution of the RF transmit field Bi strength at least on the basis of the transformed signal sets, the non-linear model of step c) preferably being adjusted accordingly.
  • Step c) may comprise calculating transformed signals derived from the acquired signals by a linear or non-linear transformation thereof.
  • the acquired signals or signals derived from the acquired signals by a linear or non-linear transformation are referred to as "measured signals", as opposed to signals that are purely calculated, for example from a model.
  • a set of measured signals corresponding to a given location in the sample is referred to as a "measured signal vector”.
  • the number of transformed signals are at least strictly larger than the number of acquisition sequences with different acquisition parameters and preferably at least equal to the number of acquired signals with different dependence on Bi .
  • step c) of the invention may comprise deriving Bi from a set of normalized signals that is derived from the acquired signals by dividing, for each imaged location in the sample, each signal corresponding to that location by the root-mean-square (RMS) value of all signals corresponding to that location, and deriving the Bi map from the normalized signals on the basis of a model of the normalized signals.
  • the measured signals may be the normalized signals.
  • Appropriate choice of the acquisition parameters of all sequences may ensure that for no Bi value in the dynamic range the RMS value of all signals is close to the measurement noise, for example lower than five times the RMS value of the measurement noise, thus avoiding high relative uncertainty in the RMS value and degrading the information content of the signal vector upon division.
  • a linear transformation of the acquired signals may comprise combining acquired signals that were obtained in response to the imaging sequences with identical acquisition parameters, for example by averaging them, in order to increase the signal-to-noise ratio.
  • a non-linear transformation of the acquired signals may comprise combining acquired signals that were obtained using a receive RF coil array, where each RF coil in the array gives rise to one set of acquired signals, for example by calculating the RMS value of the corresponding signals from all coils independently for each MR sequence.
  • the method according to the present invention allows maximizing the number of individual signals used for deriving the Bi map. This is advantageous in terms of the amount of data exploited and thus in terms of the precision of the resulting map.
  • step c) of the invention does not comprise applying a linear or non linear transformation to select part of the acquired signals and deriving Bi from the selected signals.
  • the method of the present invention may comprise a step of forming measured signal vectors on the basis of the measured signals.
  • the number of measured signal vectors may be the number of locations at which the MR signals are sampled.
  • the number of the measured signals in a measured signal vector may be the number of MR signals acquired at a given location.
  • Step c) of the invention may comprise a method to reconstruct Bi from the measured signals by non-linear fitting on the basis of a non-linear model of the measured signals as a function of at least Bi, the model consisting of a set of non- linear forward signal equations linking the true Bi to a signal set of modeled MR signals, possibly as a function of other sample parameters such as Ti, for example by optimizing model parameters to minimize a cost function representing the differences between the measured and the modeled signals, that cost function being for example the sum of squares of the absolute differences between at least part of the measured MR signals and the corresponding modeled MR signals.
  • the method may comprise subjecting, when acquiring the MR signals, the portion of the sample to time- varying magnetic field gradients for spatial localization of the signals.
  • the first and second sequences may be chosen from known Bi mapping sequences.
  • the first imaging sequence may differ from the second imaging sequence by at least one acquisition parameter, this acquisition parameter being for example chosen from echo time (TE), repetition time or any, all or a subset of the repetition times in case the sequence is characterized by more than one repetition time (TR), flip angle or any, all or a subset of flip angles in case the sequence is characterized by more than one flip angle (FA), RF pulse phase or any, all or a subset of RF pulse phases in case the sequence is characterized by more than one RF pulse phase ((PRF), RF pulse duration or any, all or a subset of RF pulse durations in case the sequence is characterized by more than one RF pulse phase (TRF), RF pulse shape or any, all or a subset of RF pulse shapes in case the sequence is characterized by more than one RF pulse shape (PS), any, all or a subset of delays between
  • TE echo time
  • TR repetition time
  • FA flip angle
  • FA flip angle
  • the first and second imaging sequences may be two stimulated echo sequences with at least one different acquisition parameter.
  • the stimulated echo sequences are for example two DREAM (dual refocusing echo acquisition mode) pulse sequences such as taught in EP 2615470, each comprising at least two preparation RF pulses followed by one or more excitation RF pulses, each of the excitation pulses giving rise to at least a gradient echo (GRE) signal and a stimulated echo (STE) signal which are refocused and acquired at different times, the sequence being defined by the acquisition parameters such as the flip angles of the preparation RF pulses, the delay between preparation RF pulses, the delay between the second preparation RF pulse and the first excitation RF pulse, the shapes and bandwidths of the preparation and excitation RF pulses, the echo times of the gradient echo and stimulated echo signals, the flip angle of the excitation RF pulses, the repetition time between successive excitation pulses, the number of excitation RF pulses, the number of phase encode steps, the amplitudes of the phase encode gradients, the
  • the parameters preferably modified between the two or more successive DREAM sequences are the preparation and excitation RF pulse flip angles.
  • the first and second imaging sequences may be two gradient echo sequences with at least one different acquisition parameter.
  • the gradient echo sequences are for example two AFI (actual flip angle imaging) pulse sequences such as taught in "Actual flip- angle imaging in the pulsed steady state: a method for rapid three-dimensional mapping of the transmitted radiofrequency field» by Yarnykh mentioned above, one such AFI sequence comprising two interleaved gradient echo pulse sequences, each acquired using excitation RF pulses with identical flip angle separated from the following excitation RF pulse by different repetition times TRi and TR 2 . Signals of the two interleaved sequences are acquired during TRi and TR 2 using gradient echoes of identical echo time.
  • AFI actual flip angle imaging
  • a single portion of k-space data is acquired for each of the sequences and acquisition is repeated until all the k-space data required to reconstruct an image at the selected field of view and resolution is acquired.
  • the parameters preferably modified between the two or more successive AFI sequences in the present invention are the excitation RF pulse flip angles and repetition times.
  • the first or second imaging sequence or both imaging sequences may be a modified AFI sequence, wherein the two interleaved gradient echo sequences are acquired using excitation RF pulses with different flip angles.
  • the first and the second imaging sequences may be of different type.
  • the first imaging sequence may be a DREAM sequence and the second imaging sequence may be an AFI sequence.
  • N MR images are generated, for example before step c), each MR image corresponding to the spatial distribution of the acquired signals Si . . . SN, giving rise, for one and preferably each imaged location x in the sample, to a signal set of N signals [SI(X). .. SN(X)] originating from that location and permitting to derive the radio-frequency transmit field strength at the corresponding location in the NMR apparatus in the presence of the sample.
  • an imaged location, a sampled location or a location at which acquired signals are sampled refer to a location of the sample from which MR signals originate and which is included in the region represented in the generated MR images.
  • the method may comprise generating MR images which sample the acquired signals at the same locations, that is, all the N MR images have the same resolution, field of view, position and angulation.
  • the method comprises generating MR images which sample the acquired signals at different locations but cover regions that overlap at least partially, reconstructing signal sets [Si(x). .. SN(X)(X)] at each location x in a set of chosen locations, where N(x) is the number of MR images that cover a region including the location x, and deriving, on the basis of at least part of the set of N(x) available signals at each location, a map of the spatial distribution of the RF transmit field strength Bi .
  • Step c) of the invention may comprise processing the measured signals, obtained by subjecting the sample to two AFI sequences characterized by acquisition parameters TRi, TR 2 , TRi ', TR 2 ', FA and FA', generating four MR images representing the spatial distribution of the acquired signals [Si, S 2 , Si ', S2'] and normalizing the signals at each image location by the RMS value of all signals at that image location, by deriving the value of at least Bi minimizing the sum-of-squares difference between the measured signals and corresponding modeled signals normalized by the RMS value of the modeled signals, the signals being for example modeled according to the known full non-linear signal equation as a function of Bi, Ti and acquisition parameters.
  • Step c) may comprise searching values of both Bi and Ti conjointly by minimizing the difference between measured and modeled signals.
  • Step c) may comprise estimating and removing common phase factors from the measured signals followed by comparing phase- corrected measured signals to the modeled signals.
  • step c) of the invention may comprise on the one hand creating a database of selected signal vectors, i.e. sets of selected signals referred to as a dictionary, which can be derived for example from a non-linear model of the measured signals for at least a series of selected strengths of RF transmit fields Bi, and on the other hand forming measured signal vectors, subsequently comparing each of the measured signal vectors to at least some of the selected signal vectors, and deriving the map of the spatial distribution of the RF transmit field Bi strength at least on the basis of such comparison.
  • the number of selected signal vectors may be the number of dictionary entries.
  • the number of signals in a selected signal vector is equal to the number of signals in a measured signal vector.
  • Step c) may comprise searching, for example for each sampled location, an element in the dictionary that corresponds best to the measured signal vectors.
  • the first acquired GRE signal in the measured signal vector is compared to the first GRE signal in the selected signal vector
  • the first STE signal in the measured signal vector is compared to the first STE signal in the selected signal vector and so on, and a single number quantifying the match between the measured and selected signal vectors is computed.
  • This may for example be the sum of the squares of the element-wise differences obtained for each of the vectors, which then needs to be minimized.
  • the measured and simulated signal vectors are normalized such that they are of norm 1. In this case, minimizing the sum of squares differences is equivalent to maximizing the scalar product between the two vectors, the latter being faster to calculate.
  • the selected signal vectors may correspond to signals simulated based on a mathematical or numerical model of the signals produced by the first and second imaging sequences, for at least selected strengths of RF transmit field Bi.
  • Selected signal vectors may also correspond to signals simulated based on a mathematical or numerical model of the signals produced by the first and second imaging sequences, for selected combinations of parameter values for at least two parameters, including the strength of the RF transmit field Biand one or more of other relevant sample and/or machine dependent parameters such as Ti, T 2 , T 2 *, Bo.
  • the dictionary may contain simulated signals corresponding to selected combinations of strengths of RF transmit field Bi and relaxation times Ti.
  • Step c) may include taking into account the uncertainty of the individual component signals based on knowledge of the noise properties of the signals acquired, for example by calculating the covariance matrix of the elements of the signal vectors using Monte-Carlo simulations and basing the comparison between signal vectors and dictionary entries on the local Mahalanobis distance between the two.
  • step c) of the invention may comprise reconstructing Bi independently for each of a set of locations of the sample by independently processing the measured signal vectors of which each comprises signals corresponding to one such location of the sample, for each such location of the sample.
  • step c) of the invention may comprise reconstructing Bi for each of a set of locations of the sample by processing the measured signal vectors of which each comprises signals corresponding to one such location of the sample, the processing performed to reconstruct Bi at one location in the sample also taking into account the measured signal vectors from other locations of the sample, for example neighboring locations. This may have the advantage of improving the speed or robustness of the processing, or both.
  • the method may comprise a step of controlling signal scale factors between sequences.
  • the MR signal is not an absolutely quantified signal, and if performing completely independent acquisitions, it is possible for data from separate sequences to each have independent scale factors. If not accounted for and unknown, such scale factors risk to cause problem to the BI reconstruction. It may be possible to infer these scale factors by analyzing signals from different voxels simultaneously.
  • scale factors can be assumed to be spatially uniform for each of the sequences, they may be inferred by performing a first reconstruction of Bi, calculating modeled signals from the reconstructed Bi using the nonlinear forward model and linearly fitting the scale factor for each sequence on the basis of the measured and modeled signals from all voxels or a subset of voxels. Measured signals can then be updated with the inferred scale factors and Bi can be estimated again.
  • the method may thus comprise inferring BI at one location based also on signals measured at other locations.
  • spatial smoothness may be exploited by considering each voxel as the center of a local neighborhood, by combining the signals from all signal sets in all voxels in that neighborhood into a large signal vector, by modeling the non-linear signal equations of all these signals as a function of the BI and spatial gradient in BI at the central voxel, and by fitting this model to the combined signal vector. While such an approach increases dimensionality and thus computational complexity, it may also prove beneficial in robustness.
  • the invention may comprise, before step a), a step of optimizing the acquisition parameters of the imaging sequences to be used by predicting the error of reconstruction of Bi based on simulations. Optimization of the sequence parameters may include the steps of
  • This optimization may take into account the dynamic range of Bi to be covered and the expected SNR of the raw MRI data.
  • the optimization may also take into account other parameters, for example to prefer sets of acquisition parameters which meet certain criteria for total acquisition time and the degree of exposition of the sample to the Bi field.
  • This optimization may further take into account the possibility of the presence of different values of Ti in the sample.
  • the present invention also provides a computer program for implementing the method of the invention, comprising codes for
  • MR magnetic resonance
  • N 2 magnetic resonance (MR) signals [Si ', S 2 ', SN 2 ']
  • the first and second imaging sequences being chosen so that at least one signal of the first and second subsets of MR signals acquired does not exhibit a cyclic dependency on Bi, or at least one signal of the first subset of MR signals and at least one signal of the second subset of MR signals have cyclic dependencies on Bi with periods that are different and not an integer multiple of each other, and
  • Reconstruction of the composite signal may comprise creating a vector of signals comprising or derived from individual signals of the acquired subsets of signals.
  • the invention makes it possible to increase dynamic range and/or provide better Signal Noise Ratio (SNR) by acquiring signals with varying cyclic dependencies on the underlying Bi .
  • SNR Signal Noise Ratio
  • the present invention also makes it possible to obtain a precision of measurement that is better than that which can be obtained from each individual acquisition in response to the respective imaging sequence. This is made possible because signals acquired with larger flip angles typically have both the highest sensitivity to Bi and the lowest dynamic range. A given signal amplitude of such a signal may then correspond to several different (potentially large) Bi values, each defined with a high precision but impossible to distinguish from each other given this signal alone. Given appropriate acquisition parameters, the simultaneous analysis of the entire signal vector provides for disambiguation and the high-precision information of otherwise ambiguous signals can thus be fully exploited.
  • Existing approaches that pick only unambiguous data based on thresholds discard this high- precision information and correspondingly loose in SNR efficiency for Bi mapping.
  • the Bi mapping approach of the invention is preferably used to characterize the RF field generated by the MR apparatus inside the sample, not primarily to characterize the relaxation properties of resonant species of the sample itself.
  • the relaxation properties of resonant species of the sample can be also measured in accordance with the present invention, if the acquisition parameters are chosen such that the signals have sufficient dependence on the relaxation properties of the sample.
  • the signal equations of signals obtained in response to the second imaging sequence do not depend on the acquisition parameters of the first imaging sequence, contrary to the signal equations of signals obtained in response to subsequent sequence blocks in the method of fingerprinting, making it possible to derive closed- form expressions of the signal equations and thus greatly facilitating the process of fitting the signal equations to the data or constructing the dictionary of signals.
  • the sample according to the present invention may be an object or being, or a part thereof, for example a human body or a part thereof, or an animal or a part thereof.
  • FIG. 1 schematically shows a MR device suitable for carrying out the methods of the invention
  • Fig. 2 illustrates the timing diagram of the imaging sequence in a standard AFI sequence as taught in the above mentioned document of Yarnykh,
  • FIG. 3 illustrates the timing diagram of the first and second imaging sequences in accordance with the invention, wherein the first and second imaging sequences are two standard AFI sequences acquired sequentially,
  • - Fig. 4 illustrates the timing diagram of the first and second imaging sequences in accordance with the invention, wherein the first and second imaging sequences are modified AFI sequences acquired sequentially
  • - Fig. 5 illustrates the dependence of the components of measured MR signal vectors on Bi under a first example parametrization of a standard AFI sequence as illustrated in Fig. 2,
  • FIG. 6 illustrates the dependence of the components of measured MR signal vectors on Bi under a second example parametrization of a standard AFI sequence as illustrated in Fig. 2,
  • Fig. 7 illustrates the dependence of the components of measured MR signal vectors on Bi under an example parametrization of two sequentially acquired AFI sequences as illustrated in Fig. 3,
  • Fig. 8 illustrates the dependence of the components of measured MR signal vectors on Bi under an example parametrization of two sequentially acquired modified AFI sequences as illustrated in Fig. 4,
  • Figs. 9 to 12 illustrate the distribution of reconstructed Bi values in the presence of measurement noise as a function of the true Bi, both relative to nominal Bi, from the four different example acquisition schemes of Figs. 5 to 8,
  • an MR device 1 With reference to Fig. 1 , an MR device 1 is shown.
  • the device comprises superconducting or resistive main magnet coils 2 such that a substantially uniform, temporally constant main magnetic field Bo is created along a z-axis through an examination volume.
  • the device further comprises a set of (1 st , 2 nd , and - where applicable - 3 rd order) shimming coils 2', wherein the current flow through the individual shimming coils of the set 2' is controllable for the purpose of minimizing Bo deviations within the examination volume.
  • a magnetic resonance generation and manipulation system applies a series of RF pulses and switched magnetic field gradients to invert or excite nuclear magnetic spins, induce magnetic resonance, refocus magnetic resonance, manipulate magnetic resonance, spatially and otherwise encode the magnetic resonance, saturate spins, and the like to perform MR imaging.
  • a gradient pulse amplifier 3 applies current pulses to selected ones of whole-body gradient coils 4, 5 and 6 along x, y and z-axes of the examination volume.
  • a digital RF frequency transmitter 7 transmits RF pulses or pulse packets, via a send-/receive switch 8, to a body RF coil 9, for example comprising parallel-transmit coils, to transmit RF pulses into the examination volume.
  • a typical MR imaging sequence is composed of a packet of RF pulse segments of short duration which taken together with each other and any applied magnetic field gradients achieve a selected manipulation of nuclear magnetic resonance.
  • the RF pulses are used to saturate, excite resonance, invert magnetization, refocus resonance, or manipulate resonance and select a portion of a body 10 positioned in the examination volume.
  • the MR signals are also picked up by the body RF coil 9.
  • a set of local array RF coils 11, 12, 13 are placed contiguous to the region selected for imaging.
  • the array coils 1 1, 12, 13 can be used to receive MR signals induced by body-coil RF transmissions.
  • the resultant MR signals are picked up by the body RF coil 9 and/or by the array RF coils 11 , 12, 13 and demodulated by a receiver 14 preferably including a preamplifier (not shown).
  • the receiver 14 is connected to the RF coils 9, 11, 12 and 13 via send-/receive switch 8.
  • a host computer 15 controls the current flow through the shimming coils 2' as well as the gradient pulse amplifier 3 and the transmitter 7 to generate any of a plurality of MR imaging sequences, such as echo planar imaging (EPI), echo volume imaging, gradient and spin echo imaging, fast spin echo imaging, and the like.
  • EPI echo planar imaging
  • the receiver 14 receives a single or a plurality of MR data lines in rapid succession following each RF excitation pulse.
  • a data acquisition system 16 performs analog-to-digital conversion of the received signals and converts each MR data line to a digital format suitable for further processing. In modern MR devices the data acquisition system 16 can be a separate computer which is specialized in acquisition of raw image data.
  • Step c) of the present invention may be performed by the reconstruction processor 17, or by an appropriately configured separate device.
  • the MR image may represent a planar slice through the patient, an array of parallel planar slices, a three-dimensional volume, or the like.
  • the image is then stored in an image memory where it may be accessed for converting slices, projections, or other portions of the image representation into appropriate format for visualization, for example via a video monitor 18 which provides a human-readable display of the resultant MR image.
  • first and second imaging sequences are two modified AFI sequences, acquired sequentially, yielding four signals, i.e. two signals in response to each of the two sequences, each comprising two interleaved acquisitions, the first sequence being acquired using two different RF pulses of respective flip angle ai and a 2 separated by the two different repetition times TRi and TR 2 and the second sequence using two different RF pulses of respective flip angles ai ' , a 2 ' separated by repetition times TRi ' and TR 2 ' .
  • the periodicity on Bi of each signal acquired becomes more complex.
  • the two signals in response to a same modified AFI sequence can exhibit different periodicity on Bi .
  • Figs. 5 to 8 Examples of acquired signals with the respective acquisition schemes Al to A4 are illustrated in Figs. 5 to 8, with a range of local RF transmit field strengths up to six times the optimized range, from 0 to 9 (Fig. 5) or 60 (Figs. 6-8) times the nominal RF transmit field strength, to better illustrate the periodic dependence of the acquired signals on Bi . In reality, only the signals obtained for relative Bi within the range for which each sequence was optimized are used in the examples of reconstruction (shaded regions).
  • a corresponding FID signal in response to each pulse, can be acquired as illustrated in Figs. 5 and 6.
  • the two FID signals are of the same periodicity as a function of relative Bi .
  • the periodicity of the signals of the first subset are different the periodicity of the signals of the second subset.
  • the signals within a same subset of signals also exhibit different periodicity.
  • Bimaps were reconstructed using a table lookup in a dictionary of simulated signals which are obtained for Bi values in the range of optimization for each example and for a range of Ti values between 100 ms and 4000 ms.
  • Figs. 9-12 show 2D histograms of reconstructed with respect to true relative Bi in the presence of noise for each of the examples. For each value of the true relative Bi, the spread of measured values in the presence of noise can thus be seen.
  • insets show a detail of the low-Bi region of the histogram (Figs. 10-12).
  • the original AFI sequence when optimized over a large dynamic range, suffers from excessive noise, especially at low Bi, due to the low flip angle required (Fig. 10).
  • Fig. 10 Low flip angle required
  • Approaches A3 and A4 perform significantly better (Figs. 11 and 12), matching or even exceeding the performance of the original AFI sequence optimized for low- dynamic-range (Fig. 9) over the entire range, for example a more than 6-fold increased range of Bi values.
  • three or more imaging sequences can be emitted and different Bi- mapping sequences can be used.

Landscapes

  • Physics & Mathematics (AREA)
  • Condensed Matter Physics & Semiconductors (AREA)
  • General Physics & Mathematics (AREA)
  • Magnetic Resonance Imaging Apparatus (AREA)

Abstract

L'invention se rapporte au domaine d'un procédé de cartographie du champ d'émission de fréquences radioélectriques B1 produit par un appareil de résonance magnétique nucléaire. La plupart des procédés actuellement disponibles pour la cartographie de B1 disposent d'une plage dynamique limitée de valeurs mesurables de B1. La présente invention permet de multiplier, par exemple d'environ dix fois, la plage dynamique observable tout en conservant une précision de mesure élevée dans toute la plage des valeurs de B1 mesurées. Des exemples de valeurs de B1 reconstruites à l'aide respectivement de procédés d'acquisition AF1 standard et de deux séquences AF1 selon l'invention illustrent que pendant que la séquence AF1 d'origine connaît un bruit excessif à un faible B1 et des effets T1 forts à un B1 élevé lorsqu'ils sont optimisés sur une grande plage dynamique, la carte B1 reconstruite selon la présente invention égale ou dépasse même les performances de la séquence AF1 d'origine optimisée pour une plage dynamique faible sur une plage augmentée de plus de 6 fois de valeurs de B1.
PCT/EP2017/073195 2016-09-14 2017-09-14 Procédé de caractérisation du champ d'émission de fréquences radioélectriques dans une nmr WO2018050777A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
EP16306167 2016-09-14
EP16306167.4 2016-09-14

Publications (1)

Publication Number Publication Date
WO2018050777A1 true WO2018050777A1 (fr) 2018-03-22

Family

ID=56985568

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/EP2017/073195 WO2018050777A1 (fr) 2016-09-14 2017-09-14 Procédé de caractérisation du champ d'émission de fréquences radioélectriques dans une nmr

Country Status (1)

Country Link
WO (1) WO2018050777A1 (fr)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20230145981A1 (en) * 2020-04-23 2023-05-11 Koninklijke Philips N.V. Mr imaging with t1 compensated b1 mapping

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2009118702A2 (fr) 2008-03-27 2009-10-01 Koninklijke Philips Electronics N.V. Imagerie par angle de bascule avec cartographie b1 améliorée pour systèmes à émissions rf multiples
EP2615470A1 (fr) 2012-01-12 2013-07-17 Koninklijke Philips Electronics N.V. Imagerie par RM avec cartographie B1
US8723518B2 (en) 2011-03-18 2014-05-13 Nicole SEIBERLICH Nuclear magnetic resonance (NMR) fingerprinting

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2009118702A2 (fr) 2008-03-27 2009-10-01 Koninklijke Philips Electronics N.V. Imagerie par angle de bascule avec cartographie b1 améliorée pour systèmes à émissions rf multiples
US8723518B2 (en) 2011-03-18 2014-05-13 Nicole SEIBERLICH Nuclear magnetic resonance (NMR) fingerprinting
EP2615470A1 (fr) 2012-01-12 2013-07-17 Koninklijke Philips Electronics N.V. Imagerie par RM avec cartographie B1

Non-Patent Citations (17)

* Cited by examiner, † Cited by third party
Title
ANTOINE LUTTI ET AL.: "Robust and fast whole brain mapping of the RF transmit field BI at 7T", PLOS ONE, vol. 7, no. 3, March 2012 (2012-03-01)
ANTOINE LUTTI ET AL: "Robust and Fast Whole Brain Mapping of the RF Transmit Field B1 at 7T", PLOS ONE, vol. 7, no. 3, 12 March 2012 (2012-03-12), pages e32379, XP055354797, DOI: 10.1371/journal.pone.0032379 *
BUONINCONTRI ET AL., MAGNETIC RESONANCE IN MEDICINE, 2015
CUNNINGHAM CH ET AL.: "Saturated Double Angle Method", MAGNETIC RESONANCE IN MEDICINE, vol. 55, 2006, pages 1326 - 1333, XP055031547, DOI: doi:10.1002/mrm.20896
F. JIRU; U. KLOSE: "Fast 3D Radiofrequency Field Mapping using Echo-Planar Imaging", MAGNETIC RESONANCE IN MEDICINE, vol. 56, 2006, pages 1375 - 1379, XP055027996, DOI: doi:10.1002/mrm.21083
FRANCESCO PADORMO ET AL: "Large dynamic range relative B1+ mapping", MAGNETIC RESONANCE IN MEDICINE., vol. 76, no. 2, 26 August 2015 (2015-08-26), US, pages 490 - 499, XP055354796, ISSN: 0740-3194, DOI: 10.1002/mrm.25884 *
GUIDO BUONINCONTRI ET AL: "MR fingerprinting with simultaneous B1 estimation", MAGNETIC RESONANCE IN MEDICINE., vol. 76, no. 4, 28 October 2015 (2015-10-28), US, pages 1127 - 1135, XP055354800, ISSN: 0740-3194, DOI: 10.1002/mrm.26009 *
HORNAK JP ET AL.: "Magnetic Field Mapping", MAGNETIC RESONANCE IN MEDICINE, vol. 6, 1988, pages 158 - 163, XP002668889, DOI: doi:10.1002/mrm.1910060204
KALINA V. JORDANOVA ET AL: "Measuring B1 distributions by B1 phase encoding", MAGNETIC RESONANCE IN MEDICINE., vol. 77, no. 1, 17 January 2016 (2016-01-17), US, pages 229 - 236, XP055354791, ISSN: 0740-3194, DOI: 10.1002/mrm.26114 *
KALINA V.J ET AL.: "Measuring Bl distributions by Bl phase encoding", MAGNETIC RESONANCE IN MEDICINE, vol. 77, pages 229 - 238
MA D ET AL.: "Magnetic resonance fingerprinting", NATURE, vol. 495, 2013, pages 187 - 192
MAGNETIC RESONANCE IN MEDICINE, vol. 71, 2013, pages 246 - 256
NEHRKE K ET AL.: "Dual Refocusing Echo Acquisition Mode", MAGNETIC RESONANCE IN MEDICINE, vol. 68, 2012, pages 1517 - 1526
PADORMO F ET AL.: "Large Dynamic Range Relative BI + Mapping", MAGNETIC RESONANCE IN MEDICINE, vol. 76, 2015, pages 490 - 499, XP055354796, DOI: doi:10.1002/mrm.25884
SACOLICK LI ET AL.: "BI Mapping by Bloch-SiegertShift", MAGNETIC RESONANCE IN MEDICINE, vol. 63, 2010, pages 1315 - 1322
T NEMTANU ET AL: "High-Dynamic-Range High-SNR B1+ Mapping Using Multiple Cyclic MR Signals", ESMRMB ANNUAL SCIENTIFIC MEETING 2016, 29 September 2016 (2016-09-29), pages 165, XP055436223 *
YARNIKH VL: "Actual Flip Angle Imaging", MAGNETIC RESONANCE IN MEDICINE, vol. 57, 2006, pages 192 - 200

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20230145981A1 (en) * 2020-04-23 2023-05-11 Koninklijke Philips N.V. Mr imaging with t1 compensated b1 mapping

Similar Documents

Publication Publication Date Title
US10627468B2 (en) Nuclear magnetic resonance (NMR) fingerprinting
US10241174B2 (en) Nuclear magnetic resonance (NMR) fingerprinting
EP2897523B1 (fr) Prise d'empreinte par résonance magnétique nucléaire (rmn)
CN105556326B (zh) 具有dixon类型的水/脂肪分离的mr成像
EP2526437B1 (fr) Procédé et système d'imagerie des propriétés électriques
EP2615470A1 (fr) Imagerie par RM avec cartographie B1
US20140070805A1 (en) Mr imaging with b1 mapping
EP2648014A1 (fr) Imagerie par RM utilisant une amélioration de contraste APT
EP2539729B1 (fr) Imagerie par résonance magnétique d'espèces chimiques au moyen d'un modèle spectral
US11373392B2 (en) System and method for magnetic resonance fingerprinting using a plurality of pulse sequence types
EP2618731B1 (fr) Imagerie d'espèces chimiques opérant par résonance magnétique
EP3084460B1 (fr) Irm d'inversion-récupération sensible à la phase avec séparation eau/graisse
US10564243B2 (en) System and method for phase cycling magnetic resonance fingerprinting (phc-MRF)
WO2016180983A1 (fr) Irm multi-échos avec correction d'erreurs de phase induite par gradient concomitant
US20140121492A1 (en) Mri with separation of different chemical species using a spectral model
JP7023954B2 (ja) プロペラｍｒイメージング
EP3497456B1 (fr) Irm avec séparation eau/graisse de type dixon
CN108431625B (zh) 具有对运动引起的扩散梯度不一致性的修正的dti
US20200033432A1 (en) Method and apparatus for identifying at least one material comprised in a voxel
WO2018050777A1 (fr) Procédé de caractérisation du champ d'émission de fréquences radioélectriques dans une nmr
EP2431760A1 (fr) Imagerie à résonance magnétique d'espèces chimiques
EP4239357A1 (fr) Mappage de diffusion par empreintes digitales par résonance magnétique

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17772639

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17772639

Country of ref document: EP

Kind code of ref document: A1