Methods and systems for in vivo and non-invasive measurement of metabolite levels

Abstract

Embodiments of a compact portable nuclear magnetic resonance (NMR) device are described, generally comprising: a housing providing magnetic shielding; an axisymmetric permanent magnet assembly disposed within the housing and having a bore, a plurality of magnetic elements that together provide well-constrained axisymmetric magnetization to generate a near-uniform magnetic dipole field B oriented along the longitudinal axis 0 A sample chamber is provided for receiving a sample, and a high magnetic permeability mild steel electrode is provided to improve the field uniformity; a shim accessory is provided, which has a coil arranged at the longitudinal axis to provide the nearly uniform magnetic field B 0 A spatial correction is provided; and a spectrometer having a control unit for measuring metabolites in a sample by applying a magnetic excitation pulse to the sample, measuring a free induction delayed signal generated by a hydrogen proton ensemble within the sample, and suppressing a water signal by using a dephasing gradient with frequency selective suppression.

Description

[0001] Cross-references

[0002] This application claims the benefit of U.S. Provisional Patent Application No. 62 / 731,576, filed on September 14, 2018, and the entire contents of U.S. Provisional Patent Application No. 62 / 731,576 are incorporated herein by reference. Technical Field

[0003] Various embodiments are described herein that relate generally to the measurement of metabolite levels, and more particularly to a portable nuclear magnetic resonance (NMR) device for in vivo and non-invasive measurement of metabolite levels. Background Art

[0004] Conventional techniques for measuring metabolite levels in blood samples are usually invasive, inaccurate and non-repeatable. For example, current blood glucose measurement techniques require piercing the skin of a subject with a lance to deposit a small amount of blood sample on a biochemically reactive, disposable and single-purpose test strip. Each blood sample requires a new deposit. The inconvenience of repeatedly piercing the skin and the cost burden associated with purchasing new test strips for each batch of samples often become obstacles to frequent glucose testing. This is particularly worrying for individuals with late stage type 1 or type 2 diabetes who need frequent and accurate testing to ensure correct insulin dosage decisions. Summary of the invention

[0005] According to the broad aspects taught herein, there is provided at least one embodiment of a method for in vivo and non-invasive quantification of glucose concentration in a sample using a portable nuclear magnetic resonance (NMR) device, the method comprising: applying a uniform static magnetic field (B0) to induce magnetization of the sample; using frequency selective suppression to suppress a water signal generated by magnetization of water located in the sample; applying a first RF excitation pulse (B1) to excite a glucose hydrogen proton ensemble; and detecting a free induction delay (FID) signal generated by relaxation of the glucose hydrogen proton ensemble.

[0006] In at least one of these embodiments, suppressing the water signal further comprises: applying a second RF excitation pulse (B1) to rotate the water magnetization to a transverse plane; and applying a dephasing pulse gradient to the water signal to dephase the coherence of its spin precession, wherein water suppression is performed so as to reduce modulation sidebands of the water signal.

[0007] In at least one of these embodiments, the static magnetic field (B0) has a magnitude between about 1.5 Tesla and about 2 Tesla.

[0008] In at least one of these embodiments, the static magnetic field (B0) has a field uniformity between about 0.01 ppm and less than about 0.1 ppm.

[0009] In at least one of these embodiments, the method includes generating the second RF excitation pulse (B1) to have a value given by the expression |sinc 2 The excitation profile is defined by (πΔfτ)|, and where 2τ is the duration of the pulse.

[0010] In at least one of these embodiments, the method includes generating the second RF excitation pulse (B1) to have an excitation profile defined by the expression |sinc(πΔfτ)| and wherein τ is the duration of the pulse.

[0011] In at least one of these embodiments, the method includes generating a signal to be transmitted at a resonant frequency (f H2O ) is the center of the second RF excitation pulse (B1), where τ is defined as τ = (f H2O -f β-glc ) -1 , and f β-glc It is the β-glucose anomer 1 The resonance frequency of the CH hydrogen group proton.

[0012] In at least one of these embodiments, the method includes generating a static magnetic field (B0) of about 1.68 Tesla, and the resonant frequency (f H2O ) is about 71.5 MHz, and τ is about 163 milliseconds.

[0013] In at least one of these embodiments, the method includes providing the second RF excitation pulse (B1) by generating an envelope modulated pulse train having a low duty cycle.

[0014] In at least one of these embodiments, the method includes generating the second RF excitation pulse (B1) using a tailored excitation for variable delay precession (DANTE) technique.

[0015] In at least one of these embodiments, the second RF excitation pulse (B1) is implemented as a hyperbolic secant pulse.

[0016] In at least one of these embodiments, the method includes increasing suppression of the water signal by generating a second excitation RF pulse (B1) of large amplitude.

[0017] In at least one of these embodiments, the dephased pulse gradient is generated by a set of DC coupled shim coils.

[0018] In at least one of these embodiments, the glucose hydrogen proton ensemble includes the α-glucose anomer 1 CH hydrogen group proton and β-glucose anomer 1 CH hydrogen group proton.

[0019] In at least one of these embodiments, the method includes generating the first RF excitation pulse (B1) such that a net magnetic moment of the glucose hydrogen proton ensemble is rotated into a transverse plane.

[0020] In at least one of these embodiments, the method comprises: in a mixture comprising the α-glucose anomer 1 CH hydrogen group proton and the β-glucose anomer 1 The first RF excitation pulse (B1) is generated in the frequency range of the Larmor frequency of the protons of the CH hydrogen radicals.

[0021] In at least one of these embodiments, the α-glucose anomer 1 The Larmor frequency of the CH hydrogen group proton is 5.223 ppm, and the anomeric isomer of β-glucose 1 The Larmor frequency of the CH hydrogen group proton is 4.634 ppm.

[0022] In at least one of these embodiments, the method includes generating the first excitation field (B1) to be left-handed circularly polarized.

[0023] In at least one of these embodiments, the method includes applying the first excitation field (B1) for less than about 1.5 ms to mitigate transverse relaxation decay of the glucose hydrogen proton ensemble at an exponential rate T2*.

[0024] In at least one of these embodiments, the method includes generating the first excitation field (B1) and the second excitation field (B1) by using canted cosine coils that are coaxially positioned relative to each other and canted in opposite directions relative to a common axis.

[0025] In at least one of these embodiments, the method further comprises: applying homonuclear decoupling to the glucose hydrogen proton ensemble, and wherein applying homonuclear decoupling comprises: 2 CH group hydrogen protons and β anomer 2 A low continuous wave radiation pulse is applied at the resonance frequency of the hydrogen protons of the CH group.

[0026] In at least one of these embodiments, the α and β anomers 2 The resonance frequencies of hydrogen protons of CH groups are 3.519 ppm and 3.23 ppm respectively.

[0027] In at least one of these embodiments, the homonuclear decoupling at least partially overlaps with applying the first RF excitation pulse (B1) to excite the glucose hydrogen proton ensemble.

[0028] In at least one of these embodiments, the method further comprises applying a discrete Fourier transform (DFT) to the FID signal to convert the FID signal into a magnetic resonance spectrum.

[0029] In at least one of these embodiments, the method further comprises: based on the resonance spectrum of the α-glucose anomer 1 CH hydrogen group proton and β-glucose anomer 1 A one-to-one mapping of the amplitudes of the resonance peaks of the CH hydrogen group protons is used to determine the glucose concentration in the sample.

[0030] In at least one of these embodiments, determining the glucose concentration in the sample comprises: 1 CH hydrogen group proton and the β-glucose anomer 1 The amplitude of the resonance peak of the CH hydrogen group protons is correlated with a known glucose concentration reference standard.

[0031] In at least one of these embodiments, determining the glucose concentration in the sample comprises: determining the alpha-glucose anomer 1 CH hydrogen group proton and the β-glucose anomer 1 The anomeric ratio of the resonance peak of the CH hydrogen group proton.

[0032] In at least one of these embodiments, the method further comprises using a non-selective inversion recovery sequence prior to applying the first RF excitation pulse (B1) to null the macromolecular response.

[0033] In at least one of these embodiments, the method further comprises using a selective inversion recovery sequence prior to applying the first RF excitation pulse (B1) to null a response of the glucose hydrogen proton ensemble and measure a macromolecular response.

[0034] In at least one of these embodiments, magnetic resonance velocimetry (MRV) technology is used to distinguish glucose molecules flowing in the blood from stationary glucose molecules.

[0035] In at least one of these embodiments, the glucose concentration is determined to have an error of less than + / - 2% at a statistical confidence level of 0.95.

[0036] According to another broad aspect of the teachings herein, at least one embodiment of a portable spectrometer for performing nuclear magnetic resonance (NMR) spectroscopy on a sample is provided, the portable spectrometer comprising: a radio frequency (RF) source configured to generate a pulsed RF signal having an in-phase component and an orthogonal component; a transmitting unit configured to receive the pulsed RF signal and generate a pulsed RF excitation field (B1), the transmitting unit comprising: a first transmission path having a first transmission bandpass filter, the first transmission bandpass filter having at least one first transmission inductor coil, the at least one first transmission inductor coil receiving the in-phase component of the RF signal and generating the in-phase component of the RF excitation field; and a second transmission path having a second transmission bandpass filter, the second transmission bandpass filter having at least one second transmission inductor coil, the at least one second transmission inductor coil receiving the orthogonal component of the RF signal and generating the orthogonal component of the RF excitation field; a receiving unit configured to receive a resonance signal generated by the sample in response to the pulsed RF excitation field (B1), the receiving unit comprising: a first receiving path , which has a first receiving bandpass filter, the first receiving bandpass filter has at least one first receiving inductive coil for receiving the in-phase component of the resonance signal; and a second receiving path, which has a second receiving bandpass filter, the second receiving bandpass filter has at least one second receiving inductive coil for receiving the orthogonal component of the resonance signal; and a processor unit, which is coupled to the RF source, the transmitting unit and the receiving unit, the processor unit is configured to: control the operation of the portable spectrometer by sending a control signal to the RF source to generate the pulsed RF signal and sending the pulsed RF signal to the transmitting unit to generate the pulsed RF excitation field (B1), and the processor unit is configured to: receive the in-phase component and the orthogonal component of the resonance signal from the receiving unit, and wherein the processor performs at least one of the following: (a) storing the in-phase component and the orthogonal component of the resonance signal in a memory unit of the processor unit for subsequent analysis; and (b) analyzing the in-phase component and the orthogonal component of the resonance signal to determine the concentration of the metabolite in the sample.

[0037] In at least one of these embodiments, each of the first and second transmit bandpass filters and the first and second receive bandpass filters is a differential bandpass filter synthesized from a tee topology low pass filter.

[0038] In at least one of these embodiments, at least one first transmit inductor coil of the first transmit bandpass filter, at least one second transmit inductor coil of the second transmit bandpass filter, at least one first receive inductor coil of the first receive bandpass filter, and at least one second receive inductor coil of the second receive bandpass filter are volume coils.

[0039] In at least one of these embodiments, each of the first transmit bandpass filter and the second transmit bandpass filter includes an outbound transmit inductor and a return transmit inductor, and each of the first receive bandpass filter and the second receive bandpass filter includes an outbound receive inductor and a return receive inductor.

[0040] In at least one of these embodiments, the transmit bandpass filter and the receive bandpass filter are at least third order bandpass filters.

[0041] In at least one of these embodiments, each of the at least one first transmit inductive coil of the first transmit bandpass filter and the at least one second transmit inductive coil of the second transmit bandpass filter is a tilted cosine coil positioned coaxially relative to each other and tilted in opposite directions relative to a common axis so as to generate an RF excitation field (B1) that is at least one of circular polarization and elliptical polarization.

[0042] In at least one of these embodiments, each of the at least one first receive inductor coil of the first receive bandpass filter and the at least one second receive inductor coil of the second receive bandpass filter is a tilted cosine coil coaxially positioned relative to each other and tilted in opposite directions relative to a common axis.

[0043] In at least one of these embodiments, the common axis is orthogonal to an axis of the static magnetic field (B0).

[0044] In at least one of these embodiments, the resonance signal is a free induction delay (FID) signal generated by decay of magnetization in the transverse plane.

[0045] In at least one of these embodiments, the receiving unit uses a sum and a difference of the FID signals received by the two tilted cosine coils to distinguish projections of the magnetization on the two transverse spatial axes.

[0046] In at least one of these embodiments, each of the first receive path and the second receive path includes: a transformer including a primary winding coupled to an output of the receive bandpass filter, wherein the transformer is configured to provide galvanic isolation, impedance matching, and common-mode noise suppression; a low noise amplifier (LNA) coupled to a first node of a secondary winding of the transformer, wherein the LNA is configured to achieve a low noise figure and provide a uniform spectral noise distribution in a filtered resonant signal; a variable gain amplifier (VGA) coupled to an output of the low noise amplifier, wherein the VGA is configured to enhance the filtered resonant signal and minimize overall receiver noise; a local oscillator (LO) coupled to an output of the variable gain amplifier, wherein the LO is configured to generate an intermediate frequency; and an analog-to-digital converter (ADC) coupled to an output of the LO.

[0047] In at least one of these embodiments, the second node of the secondary winding of the transformer is center-tapped for local ground reference, and wherein the turns ratio of the transformer is selected for optimal impedance matching.

[0048] In at least one of these embodiments, the transformer comprises a wideband transformer balun.

[0049] In at least one of these embodiments, the LNA is formed from GaAs E-pHEMT technology and is configured to provide a noise figure below 1 dB.

[0050] In at least one of these embodiments, the VGA includes an automatic gain controller (AGC) configured to automatically reduce a gain of the VGA to prevent overload.

[0051] In at least one of these embodiments, the local oscillator is configured for a frequency deviation of at least 100 kHz.

[0052] In at least one of these embodiments, the noise generated by the receiving unit is less than 1.1 nV / √Hz (referenced to input (RTI)).

[0053] In at least one of these embodiments, the processor unit includes at least one of a digital signal processor (DSP) and a field programmable gate array (FPGA).

[0054] In at least one of these embodiments, the first transmit bandpass filter and the second transmit bandpass filter are configured to pass a first frequency passband range including at least one isotopic Larmor frequency.

[0055] In at least one of these embodiments, the first frequency passband range includes two isotope Larmor frequencies, and the transmit bandpass filter is configured to support heteronuclear measurements.

[0056] In at least one of these embodiments, the two isotope Larmor frequencies are related to fluorine ( 19 F) and hydrogen ( 1 H) is related to the Larmor frequency.

[0057] In at least one of these embodiments, the first frequency passband ranges between 60 MHz and 80 MHz.

[0058] In at least one of these embodiments, the first receive bandpass filter and the second receive bandpass filter are configured to pass a second frequency passband range including at least one isotope Larmor frequency.

[0059] According to another broad aspect of the teachings herein, a method for performing nuclear magnetic resonance (NMR) spectroscopy on a sample using a portable spectrometer is provided, the method comprising: applying a static magnetic field (B0) to the sample; using a radio frequency (RF) source to generate a pulsed RF signal having an in-phase component and an orthogonal component; sending the in-phase component of the pulsed RF signal to a first transmit bandpass filter of a transmit unit to generate a filtered in-phase RF component; sending the orthogonal component of the pulsed RF signal through a second transmit bandpass filter of the transmit unit to generate a filtered orthogonal RF component, wherein the first transmit bandpass filter and the second transmit bandpass filter have a first bandpass range, the first bandpass range including at least one Larmor frequency of a metabolite to be measured; applying the filtered in-phase RF component to at least one first transmit inductive coil of the first transmit bandpass filter and applying the filtered orthogonal RF component to The method comprises: generating a pulsed RF excitation field (B1) by using at least one second transmitting inductive coil of the second transmitting bandpass filter; applying the pulsed RF excitation field (B1) to the sample; receiving an in-phase component of a resonance signal generated by the sample using at least one first receiving inductive coil of a first receiving bandpass filter of a receiving unit; receiving a quadrature component of the resonance signal generated by the sample using at least one second receiving inductive coil of a second receiving bandpass filter of the receiving unit, wherein the first receiving bandpass filter and the second receiving bandpass filter have a second passband range, and the second passband range includes at least one Larmor frequency of a metabolite to be measured; and passing the in-phase component of the resonance signal through the first receiving bandpass filter to generate a filtered in-phase component of the resonance signal, and passing the quadrature component of the resonance signal through the second receiving bandpass filter to generate a filtered quadrature component of the resonance signal.

[0060] In at least one of these embodiments, the resonance signal is a free induction delay (FID) signal generated by decay of magnetization in the transverse plane.

[0061] In at least one of these embodiments, the method further includes passing each of the in-phase component and the quadrature component of the resonance signal through a low noise amplifier, a variable gain amplifier, a local oscillator, and an analog-to-digital converter (ADC).

[0062] According to another broad aspect of the teachings herein, there is provided a compact magnet assembly for generating a uniform static magnetic field (B0) across a bore of a portable nuclear magnetic resonance (NMR) device, the compact magnet assembly comprising: an axisymmetric segmented permanent magnet assembly rotated about an axis of symmetry (z-axis), the permanent magnet being configured to generate a near-space uniform static magnetic field (B0) across the bore, wherein the permanent magnet assembly comprises: a top disk-conical magnet segment stacked above the bore in a +z direction and having a magnetization in the +z direction; a bottom disk-conical magnet segment stacked below the bore in a -z direction and having a magnetization in the +z direction; and a central annular magnet segment positioned radially outward (r) from the bore in a +r direction and having a magnetization in the -z direction, wherein the central annular magnet at least partially surrounding the hole while maintaining an unobstructed entrance to the hole; a top annular magnet segment stacked above the center annular segment in the +z direction and arranged radially outward from the top disk-conical magnet segment in the +r direction, wherein the top annular magnet segment has a magnetization in the +r direction; and a bottom annular magnet segment stacked below the center annular segment in the -z direction and arranged radially outward from the bottom disk-conical magnet segment in the +r direction, wherein the top annular magnet segment has a magnetization in the -r direction, wherein the superposition of the magnetic fields generated by each magnet segment generates a near-spatial uniform static magnetic field (B0) across the hole along the symmetry axis; and a shim device configured to provide spatial magnetic field correction to the near-spatial uniform static magnetic field.

[0063] In at least one of these embodiments, the uniform static magnetic field (B0) generated across the magnet bore is greater than about 1.5 Tesla.

[0064] In at least one of these embodiments, the uniform static magnetic field (B0) generated across the aperture is between about 1.5 Tesla and about 2 Tesla.

[0065] In at least one of these embodiments, the uniform static magnetic field (B0) has a field uniformity of less than about 0.1 ppm.

[0066] In at least one of these embodiments, the static magnetic field (B0) has a field uniformity of substantially 0.01 ppm.

[0067] In at least one of these embodiments, the permanent magnet is characterized by a three-dimensional confinement of 5 Gauss field lines.

[0068] In at least one of these embodiments, each magnet segment of the permanent magnet assembly is formed of a hard permanent magnet alloy that allows the magnetic fields generated by each magnet segment in the permanent magnet assembly to superimpose.

[0069] In at least one of these embodiments, the hard permanent magnet is neodymium (NdFeB).

[0070] In at least one of these embodiments, the center ring magnet segment is formed of N40 grade NdFeB to counteract high reverse coercive fields, and wherein the top and bottom ring magnet segments are formed of N40 grade NdFeB to provide optimal field confinement.

[0071] In at least one of these embodiments, an inner surface of the central annular magnet segment facing the bore includes a correction magnet segment that is curved radially inward and configured to improve uniformity of a close-spaced uniform static magnetic field (B0) generated across the bore.

[0072] In at least one of these embodiments, the correction magnet segments are pole pieces formed from high permeability mild steel.

[0073] In at least one of these embodiments, the shimming device is powered by a lithium-ion (Li-Ion) battery.

[0074] In at least one of these embodiments, the portable nuclear magnetic resonance (NMR) device is provided in a compact and portable form suitable for a home environment and suitable for encouraging diabetic patients and other patients with other metabolite disorders to use the device more frequently, thereby facilitating frequent glucose or other metabolite testing.

[0075] According to another broad aspect of the teachings herein, there is provided a shimming device configured to provide spatial magnetic field correction to a near-space uniform static magnetic field, wherein the shimming device comprises: a plurality of linear current-carrying conductors arranged in a circular configuration, and wherein the plurality of linear current-carrying conductors are evenly spaced around the circumference of the circular configuration.

[0076] In at least one of these embodiments, each of the plurality of linear current carrying conductors is driven by a respective DC current, and wherein each of the plurality of linear current carrying conductors has a uniform density.

[0077] In at least one of these embodiments, the current distribution of the respective DC currents of the plurality of linear current carrying conductors is a sinusoidal distribution of DC current that varies as a function of the angular position of the linear current carrying conductors around the circumference of the circular configuration, and wherein the shimming device is configured to generate a high-order shimming pattern.

[0078] In at least one of these embodiments, the near-space uniform static magnetic field is represented as a first nth-order polynomial having n first coefficients, and the plurality of linear load-carrying conductors include m linear current-carrying conductors, wherein each of the m linear current-carrying conductors carries a current having an amplitude that is a superposition of n current modes, wherein each current mode primarily corresponds to a term in the first nth-order polynomial.

[0079] In at least one of these embodiments, the spatial magnetic field correction generated by the field shim device is represented as a second nth-order polynomial having n second coefficients, wherein the n second coefficients correspond to the amplitude of the modal current required to produce the desired compensated nth-order static field (B0) field polynomial, wherein there is a linear relationship between the n first coefficients and the n second coefficients.

[0080] According to another broad aspect taught herein, there is provided a shim device configured to provide spatial magnetic field correction for a static magnetic field having a nearly spatially uniform profile, wherein the shim device comprises a first set and a second set of concentric current-carrying conductors, wherein: each concentric current-carrying conductor in the first set of concentric current-carrying conductors is spaced apart and arranged oppositely from a corresponding concentric current-carrying conductor in the second set of current-carrying conductors along an axis of the static magnetic field to form a plurality of current-carrying conductor pairs, wherein each current-carrying conductor generates a compensating magnetic field to correct the nearly spatially uniform profile of the static magnetic field.

[0081] In at least one of these embodiments, each of the plurality of current carrying conductors is driven by a respective DC current.

[0082] In at least one of these embodiments, at least one current-carrying conductor pairing is configured such that a coil radius of each coil in the current-carrying conductor pairing is substantially equal to a spacing between coils of the conductor pairing, thereby satisfying a Helmholtz condition, and the conductor pairing generates a linear field gradient between the coils of the conductor pairing.

[0083] In at least one of these embodiments, the plurality of current carrying conductors are coupled to a DC-DC converter configured to boost current from a DC power source to provide current to each of the plurality of current carrying conductors, whereby the DC-DC converter allows for varying shimming capabilities of each current carrying conductor so that the shimming device provides greater spatial magnetic field correction to the static magnetic field.

[0084] In at least one of these embodiments, the current carrying conductors in each of the first and second sets are offset from one another, and the offset is within a range between approximately 10 and 45 degrees offset.

[0085] In at least one of these embodiments, each of the first and second sets of concentric current carrying conductors are arranged in a disk configuration, and the disk configuration has a disk radius, and wherein a diameter of each current carrying conductor is less than the disk radius.

[0086] In at least one of these embodiments, each of the first and second sets of concentric current carrying conductors are arranged in a disk configuration, and the disk configuration has a disk radius, and wherein a diameter of each current carrying conductor is approximately equal to the disk radius.

[0087] In at least one of these embodiments, the shimming device is configured to generate a spatial magnetic field correction of the static magnetic field that varies with azimuth.

[0088] In at least one of these embodiments, each of the first and second sets of current carrying conductors includes a plurality of nested current carrying conductors, each current carrying conductor having a center point that is co-linear with an axis of the static magnetic field (B0).

[0089] In at least one of these embodiments, the magnetic field generated by each current carrying conductor is represented using a polynomial expansion of spherical harmonics, and wherein each current carrying conductor generates even and odd polynomial terms.

[0090] In at least one of these embodiments, current carrying conductors having different radii generate linearly independent polynomial coefficient vectors that allow any desired compensation field with azimuthal symmetry to be generated within the volume of a magnet bore of an axisymmetric permanent magnet assembly that generates the static magnetic field in the magnet bore.

[0091] According to another broad aspect taught herein, there is provided a method for in vivo and non-invasive quantification of the concentration of small molecule metabolites in a sample using a portable nuclear magnetic resonance (NMR) device, the method comprising: selecting a high-resolution resonance feature subset from a plurality of resonance features associated with the metabolites; applying a uniform static magnetic field (B0) to induce magnetization of the sample; applying a first RF excitation pulse (B1) to the sample, wherein the first RF excitation pulse (B1) is applied at a frequency range including at least one Larmor frequency associated with resonance features of the high-resolution resonance feature subset; and detecting a free induction delay (FID) signal generated by the sample.

[0092] In at least one of these embodiments, the small molecule metabolite comprises at least one of: glucose, glycogen, BHB, and a ketoacidosis marker.

[0093] In at least one of these embodiments, magnetic resonance velocimetry (MRV) technology is used to distinguish small molecule metabolites circulating in the blood from stationary small molecule metabolites.

[0094] In at least one of these embodiments, the high resolution resonant feature subset is characterized by a high signal-to-noise ratio (SNR).

[0095] In at least one of these embodiments, the associated plurality of resonance features comprises a plurality of chemical shift resonances associated with the metabolite.

[0096] According to another broad aspect taught herein, a compact portable NMR device is provided, comprising: a housing providing magnetic shielding; a permanent magnet accessory disposed within the housing and having a hole along a portion of the longitudinal axis of the permanent magnet accessory, the permanent magnet accessory having a plurality of magnetic elements that together provide an axisymmetric magnetization to generate a near-uniform magnetic field B0 oriented along the longitudinal axis, and one of the magnet elements having a sample cavity sized to receive a sample; a hollow frame disposed between the permanent magnet accessory and the housing and providing an internal space; a shim accessory having a coil disposed at the longitudinal axis to provide spatial magnetic field correction to the near-uniform magnetic field B0; and a spectrometer disposed within the hollow space of the frame and coupled to the magnetic hole, the spectrometer having a control unit for applying magnetic excitation pulses to the sample and measuring a free induction delay signal generated by a hydrogen proton ensemble within the sample.

[0097] In at least one of these embodiments, the housing is sized to be held in a tabletop stand and the sample cavity is sized to receive a sample insert that holds the sample.

[0098] In at least one of these embodiments, the housing is sized to be held by a hand of a subject, the sample is provided by a finger or thumb of the subject, and the sample cavity is sized to receive a finger of the subject.

[0099] In at least one of these embodiments, the device further comprises a sliding door to selectively allow access to the sample chamber during measurement.

[0100] In at least one of these embodiments, in at least one of these embodiments, the device further includes a touch sensor that can be touched to activate the device.

[0101] In at least one of these embodiments, the device further comprises a display for providing a user interface and displaying measurement results.

[0102] In at least one of these embodiments, the apparatus further comprises a communication module for allowing the control unit to communicate with a remote device and receive control data and measurements.

[0103] In at least one of these embodiments, the control unit is configured to perform a method described in accordance with the teachings herein, including but not limited to one of: a method for in vivo and non-invasive quantification of glucose concentration in a sample using a portable NMR device, a method for in vivo and non-invasive quantification of the concentration of small molecule metabolites in a sample using a portable NMR device, or a method for performing NMR spectroscopy on a sample using a portable spectrometer.

[0104] In at least one of these embodiments, the spectrometer, permanent magnet assembly, and shimming device are defined according to the teachings herein.

[0105] Other features and advantages of the present application will become apparent through the following detailed description in conjunction with the accompanying drawings. However, it should be understood that the detailed description and specific examples, although indicating the preferred embodiments of the present application, are only for illustrative purposes, because various changes and modifications within the spirit and scope of the present application will be apparent to those skilled in the art based on the detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0106] To better understand the various embodiments described herein, and to more clearly show how these various embodiments may be implemented, reference will be made to the accompanying drawings, which illustrate at least one example embodiment, by way of example, and are now described. The drawings are not intended to limit the scope of the teachings described herein.

[0107] Figure 1A An example embodiment of an in vivo and non-invasive metabolite testing system including a portable nuclear magnetic resonance (NMR) device is shown.

[0108] Figure 1B An exemplary embodiment of an in vitro metabolite testing system is shown that includes an NMR instrument suitable for a laboratory or institutional setting.

[0109] Figure 2A An exemplary hydrogen ion after exposure to a static magnetic field (B0) is shown. 1 H proton.

[0110] Figure 2B The results show that after exposure to a radio frequency (RF) excitation magnetic field (B1), Figure 2A Exemplary hydrogen 1 H protons, the magnetic field will1 The magnetization of the H protons rotates into the transverse plane.

[0111] Figure 2C and 2D shows the relaxation back to equilibrium after removal of the RF excitation magnetic field (B1). Figure 2A Exemplary hydrogen 1 H proton.

[0112] Figure 2E A simplified illustrative model of T1 and T2 relaxation is shown.

[0113] Figure 2F Shown is a graph of an exemplary decaying free induction delay (FID) signal due to T2 relaxation.

[0114] Figure 3A An exemplary nuclear magnetic resonance (NMR) graph is shown, illustrating various 1 The Larmor frequency of H hydrogen protons, taking into account each 1 H Chemical shift properties of hydrogen protons.

[0115] Figure 3B shows that only the α-anomer of glucose is used 1 Plot of statistical confidence levels for glucose concentration measurements of resonance peaks generated by protonation of CH hydrogen groups.

[0116] Figure 3C shows the use of α and β anomers of glucose 1 A graph showing the statistical confidence level of glucose concentration measurements for the resonance peaks generated by both CH hydrogen group protons.

[0117] Figure 3D A graph showing an exemplary amplitude frequency response after applying a frequency selective pulse for water signal suppression is shown.

[0118] Figure 3E A diagram showing the effect of an exemplary pulsed field gradient is shown.

[0119] Figure 3F An exemplary bipolar gradient field used during magnetic resonance velocimetry (MRV) is shown.

[0120] Figure 3G The results for protons with fixed spin and protons moving at a constant speed are shown. Figure 3F The effect of bipolar gradient field.

[0121] Figure 3H The results show that it can be applied to α- and β-glucose anomers. 1 An example embodiment of a series of magnetic signals for measuring blood glucose concentration using CH hydrogen radical proton resonance.

[0122] Figure 4A The present invention shows at least one embodiment described in accordance with the teachings of this document. Figure 1A Simplified block diagram of a portable NMR device.

[0123] Figure 4B yes Figure 1A A simplified circuit diagram of an example embodiment of a portable NMR device.

[0124] Figure 5A is Figure 1A Circuit diagram of an example embodiment of a transmission circuit for use in a portable NMR device.

[0125] Figure 5B is Figure 1A Circuit diagram of an example embodiment of a receiver circuit for use in a portable NMR device.

[0126] Figure 5C Shown in Figure 5A In the transmission circuit or Figure 5B An example frequency response of a bandpass filter used in a receiver circuit.

[0127] Figure 6 FIG. 1 shows a method for performing a multi-function circuit according to at least one embodiment of the present teachings. Figure 1A Schematic diagram of an example coil assembly in a portable NMR device.

[0128] Fig. 7A Shown along Figure 1A 7 ′ and illustrates a permanent magnet assembly within the portable NMR device according to at least one embodiment in accordance with the teachings herein.

[0129] Figure 7B Shows Fig. 7A Exploded view of the permanent magnet assembly.

[0130] Figure 7C The simulation shows that Fig. 7A and 7B A contour field plot of the magnetic field components generated by a permanent magnet assembly.

[0131] Fig.7D shows the radial distance from the magnet isocenter as a function of the Fig. 7A and 7B A graph showing the strength of the static magnetic field (B0) generated by a permanent magnet assembly.

[0132] Fig. 7E Shown by Fig. 7A and 7BGraph of a magnetostatic simulation of the static magnetic field (B0) generated by a permanent magnet assembly.

[0133] Figure 7F Shown along Figure 1A 7-7' of a portable NMR device and illustrates a permanent magnet assembly within the portable NMR device according to another example embodiment in accordance with the teachings herein.

[0134] Figure 7G The simulation shows that Figure 7F A contour field plot of the magnetic field components generated by a permanent magnet assembly.

[0135] Figure 7H shows the radial distance from the magnet isocenter as a function of the Figure 7F A graph showing the strength of the static magnetic field (B0) generated by a permanent magnet assembly.

[0136] Fig. 8A An example shim assembly is shown according to at least one embodiment in accordance with the teachings herein.

[0137] Figure 8B Shown along Fig. 8A Section line 8B-8B Fig. 8A A cross-sectional view of a shim assembly.

[0138] Figure 8C Shown along Fig. 8A Section line 8C-8C Fig. 8A Another cross-sectional view of the shim assembly.

[0139] Fig.8D Another example shim assembly according to at least one embodiment in accordance with the teachings herein is shown.

[0140] Fig. 8E shows that the shim current is applied to Fig.8D Example graphs of simulation results comparing the profile of the static magnetic field (B0) inside the bore of an axisymmetric magnet array before shimming assembly and after applying a first iteration of the method to adjust the current in each shim coil to compensate for static magnetic field inhomogeneities.

[0141] Figure 8F shows the coverage of the high-order polynomial fit Fig. 8E Example plot of a zoomed-in view of the first iteration shimming results.

[0142] Figure 8G Shown by Figure 8F A curve graph of the polynomial fitting error obtained by polynomial fitting.

[0143] Figure 8HAccording to at least one embodiment of the teachings herein, Fig.8D Example construction of a shim component.

[0144] Figure 8I According to at least another embodiment of the present invention, Fig.8D Another example configuration of a shim assembly.

[0145] Figure 8J Another embodiment according to the teachings of this document is shown. Fig.8D Yet another example configuration of a shim assembly.

[0146] Fig. 9 An example block diagram of a power and control system for a portable NMR device is shown.

[0147] Fig.10 It is used for α and β glucose anomers 1 Flow chart of an example embodiment of a method for measuring blood glucose concentration by CH hydrogen group proton resonance.

[0148] Fig.11 It is used to adjust Fig.8D Flowchart of an example method for controlling currents in shim coils in a shim assembly to correct for static magnetic field (B0) inhomogeneities in an axisymmetric magnetic bore.

[0149] Other aspects and features of the example embodiments described herein will become apparent from the following description taken in conjunction with the accompanying drawings. DETAILED DESCRIPTION

[0150] Various embodiments according to the teachings of this article will be described below to provide an example of at least one embodiment of the subject matter claimed for protection. The embodiments described herein do not limit any subject matter claimed for protection. The subject matter claimed for protection is not limited to the device, system or method with all the features of any device, system or method described below, nor is it limited to the features common to multiple or all devices, systems or methods described herein. It is possible that the device, system or method described herein is not an embodiment of any subject matter claimed for protection. Any subject matter described herein that is not claimed for protection in the file may be the subject matter of another protective instrument, such as a continuing patent application, and the applicant, inventor or owner has no intention of abandoning, abandoning or committing to disclosing any such subject matter through the disclosure in this document.

[0151] It will be understood that for simplicity and clarity of illustration, reference numerals may be repeated between the drawings to indicate corresponding or similar elements or steps, where deemed appropriate. In addition, many specific details are set forth in order to provide a thorough understanding of the example embodiments described herein. However, it will be understood by those of ordinary skill in the art that the embodiments described herein may be practiced without these specific details. In other cases, well-known methods, processes, and components are not described in detail so as not to make the embodiments described herein unclear. Moreover, this description should not be considered as a limitation on the scope of the example embodiments described herein.

[0152] It should also be noted that the terms "coupled" or "coupled" as used herein can have several different meanings, depending on the context in which these terms are used. For example, the terms "coupled" or "coupled" can have mechanical, fluid, or electrical meanings. For example, as used herein, the terms "coupled" or "coupled" can mean that two elements or devices can be directly connected to each other, or can be connected via one or more intermediate elements or devices via electrical or magnetic signals, electrical connections, electrical components, or mechanical components, depending on the specific context. In addition, the coupled electronic components can send and / or receive data.

[0153] Unless the context requires otherwise, throughout the following specification and claims, the word "comprise" and variations thereof (such as "comprises" and "comprising") are to be interpreted in an open, inclusive sense, ie, "including, but not limited to."

[0154] It should also be noted that, as used herein, the term "and / or" is intended to mean inclusive or. That is, for example, "X and / or Y" is intended to mean X or Y or both. As another example, "X, Y and / or Z" is intended to mean X or Y or Z or any combination thereof.

[0155] It should be noted that terms of degree such as "substantially", "approximately" and "approximately" as used herein refer to a reasonable amount of deviation of the modified term such that the end result is not significantly changed. These terms of degree can also be construed as including deviations of the modified term, such as 1%, 2%, 5% or 10%, if the deviation does not negate the meaning of the term it modifies.

[0156] In addition, the recitation of numerical ranges with endpoints herein includes all numbers and fractions assumed within that range (e.g., 1 to 5 includes 1, 1.5, 2, 2.75, 3, 3.90, 4, and 5). It should also be understood that all numbers and fractions thereof are assumed to be modified by the term "about", which means that up to a certain amount of variation, such as 1%, 2%, 5%, or 10%, can be referred to if the end result does not vary significantly.

[0157] Throughout the specification, references to "one embodiment," "an embodiment," "at least one embodiment," or "some embodiments" mean that one or more particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments, unless otherwise specified as not combinable or as alternative options.

[0158] As used in this specification and the appended claims, the singular forms "a," "an," and "the" include plural referents unless the content clearly dictates otherwise. It should also be noted that the term "or" is generally used in its broadest sense, meaning "and / or," unless the content clearly dictates otherwise.

[0159] The title and abstract of the disclosure provided herein are for convenience only and do not interpret the scope or meaning of the embodiments.

[0160] Similarly, throughout the specification and the appended claims, the term "communication", such as "communication path", "communicative coupling", and variations such as "communicatively coupled", is generally used to refer to any design for transmitting and / or exchanging information. Examples of communication paths include, but are not limited to, conductive paths (e.g., conductive wires, conductive traces), magnetic paths (e.g., magnetic media), optical paths (e.g., optical fibers), electromagnetic radiation paths (e.g., radio waves), or any combination thereof. Examples of communicative couplings include, but are not limited to, electrical couplings, magnetic couplings, optical couplings, radio couplings, or any combination thereof.

[0161] Throughout the specification and the appended claims, infinitive verb forms are frequently used. Examples include, but are not limited to: "detect," "provide," "send," "communicate," "process," "route," etc. Unless the specific context requires otherwise, such infinitive verb forms are to be used in an open, inclusive sense, i.e., "at least detect," "at least provide," "at least send," etc.

[0162] Example embodiments of the systems and methods described herein may be implemented as a combination of hardware or software. In some cases, the devices may be implemented, at least in part, by using one or more computer programs, in a device including at least one processing element and a data storage element (including volatile memory, non-volatile memory, storage element, or any combination thereof). The devices may also have at least one input device (e.g., keyboard, mouse, touch screen, etc.), and at least one output device (e.g., display screen, printer, radio, etc.), depending on the nature of the device.

[0163] As described in the background section, conventional techniques for measuring metabolite levels in blood samples are often invasive, inaccurate, and non-repeatable. For example, current blood glucose measurement techniques require puncturing the subject's skin with a puncture device to deposit a small amount of blood sample on a biochemically reactive, disposable, single-purpose test strip. A new deposit is required for each blood sample. The inconvenience of repeatedly puncturing the skin and the cost burden associated with purchasing new test strips for each batch of samples often become obstacles to frequent blood glucose testing. This is particularly worrying for individuals with type 1 diabetes or advanced stage type 2 diabetes who require frequent testing to ensure correct insulin dosage decisions.

[0164] Current glucose testing technology also suffers from measurement inaccuracies. On average, glucose measurement test strips have an error range of between 5% and 20% at a statistical confidence level of 0.95. This inaccuracy is typically caused by manufacturing tolerances as well as external influences that affect the electrochemical reaction of the test strip (e.g., temperature, humidity, altitude, hematocrit level, and the presence of common medications). It is believed that the measurement error of a glucose meter needs to be less than + / - 2% at a statistical confidence level of 0.95 to sufficiently reduce errors in insulin dosing decisions and thereby avoid the progression of diabetes and an increased risk of diabetes-related diseases (e.g., heart disease, kidney disease, and other diabetes-related diseases) (see, e.g., JC Boyd and DE Bruns [1]).

[0165] Sharing between users of blood glucose testing lancets used to pierce the skin has also been associated with infection. An outbreak of hepatitis B in a healthcare setting has been traced to shared test lancets, and unsafe practices have been reported at public health fairs that provide blood glucose testing services.

[0166] Attempts to develop alternative non-invasive glucose meters have been largely unsuccessful and otherwise clinically infeasible. For example, some attempts have used optical measurements including Raman or near-infrared spectroscopy to detect the concentration of glucose in circulating blood. However, since light cannot penetrate the skin deeply, these attempts are largely ineffective. In addition, reflected or transmitted light (i.e., light used to measure glucose concentration) generally lacks sufficient spectral resolution to distinguish glucose from many glucose-like molecules (e.g., glycated proteins) in the human body. Other attempts that rely on microwaves and thermal spectroscopy are also ineffective for similar reasons.

[0167] In the embodiments described herein, a nuclear magnetic resonance (NMR) device (also referred to as magnetic resonance spectroscopy (MRS) device) is provided that can be used for repeated, non-invasive and non-destructive testing of various common metabolites in a subject or test sample. The ability of the NMR device to measure and quantify various metabolites allows the device to be widely used in the management of various metabolic diseases (e.g., diabetes) and clinical research.

[0168] In various embodiments, the NMR device can be provided as a portable device that can be available in a convenient, compact and easily transportable form and suitable for everyday consumer use. The portable device can perform repeated in vivo or in vitro metabolite measurements by scanning a subject's finger or test blood sample received in a side hole (or cavity) of the device. In particular, by scanning a subject's finger (rather than puncturing the skin), and eliminating the need to purchase a new single-use test strip for each measurement, patients are encouraged to frequently use the portable NMR device for metabolite level monitoring.

[0169] In one example application, the portable NMR device was configured to perform a glucose test and return a glucose concentration result with a measurement uncertainty of less than + / - 2% at a statistical confidence level of 0.95. This level of accuracy ensures that errors in insulin dosing decisions are minimized and improves glycemic control for diabetic patients.

[0170] In at least some embodiments, the NMR device is configured to provide highly accurate metabolite level measurements using only one or two scans of a subject's finger or test sample and provide results in a competitive time of 5 to 12 seconds.

[0171] In particular, and as described in more detail herein, the portable NMR device is operable to provide high measurement accuracy by employing a novel and compact axisymmetric segmented permanent magnet assembly that applies a strong and nearly spatially uniform static magnetic field across a hole (or cavity) that receives a subject's finger or test sample. The high magnetic field strength produced by the compact permanent magnet assembly allows the device to generate metabolite level measurements with a high signal-to-noise ratio (SNR). In at least one embodiment, the permanent magnet assembly can produce a static magnetic field strength greater than 1.5 Tesla but less than 2 Tesla (in order to comply with various consumer safety regulations). In at least some embodiments, the strength of the static magnetic field produced by the permanent magnet assembly is at least 1.68 Tesla, which represents an improvement over current compact permanent magnet assemblies used in NMR applications.

[0172] In the embodiments described herein, the nearly uniform static field produced by the permanent magnet assembly is corrected by a shimming assembly, which corrects the uniformity of the static field to less than 0.1ppm (parts per million). In at least some cases, the shimming assembly can correct the uniformity of the static field to about 0.01ppm. This level of field uniformity provides high spectral resolution and allows a variety of metabolites to be quantified with a high degree of statistical confidence. In various embodiments described herein, the shimming assembly requires an accurately generated current to achieve a target uniformity of about 0.01ppm by correcting the spatial inhomogeneity of the static field. Compared with the traditional Halbach magnet design (having a similar working volume in the magnet bore), the axisymmetric segmented permanent magnet design requires less shimming current, because its inherent B0 field uniformity is higher. In at least some cases, the lower current demand for the shimming assembly allows the shimming assembly to be driven by a small power supply that can be set inside the portable NMR device.

[0173] The proposed axisymmetric permanent magnet assembly also exhibits improved three-dimensional magnetic field confinement properties. More particularly, the permanent magnet is characterized by confining 5 Gauss magnetic field lines and is otherwise compatible with every day consumer applications.

[0174] Previous attempts to develop portable NMR devices have faced challenges in developing permanent magnet architectures that similarly generate sufficient static field strength and field uniformity to produce measurements with sufficient accuracy while satisfying field constraint rules. For example, large and complex magnet designs are often used to generate strong magnetic fields, but are incompatible with local and clinical uses. Where smaller, more compact permanent magnet assemblies have been used (e.g., Halbach cylindrical magnet designs), these magnet assemblies produce static fields with poor field uniformity, which requires strong shimming assemblies whose current requirements cannot be provided by small portable power supplies.

[0175] In various embodiments also described herein, a portable NMR device uses a transmit coil and a receive coil to generate and receive an RF excitation field (B1). These coils are incorporated into a passive bandpass filter. The passive bandpass filter avoids the use of adjustable circuit elements that generate noise and otherwise affect measurement accuracy. The bandpass filter can also allow for simplified control and filtering of transmitted and received frequency signals.

[0176] According to the teachings provided herein, the portable NMR device also uses a spectrometer receiving unit that generates a signal with a low signal-to-noise ratio (SNR). In various cases, the receiving unit has a low noise design, which is characterized by a noise of less than 1.1nV / √Hz (as reference input (RTI)). By using a combination of a low noise amplifier, a local oscillator, an automatic gain controller and a high resolution analog-to-digital converter, the receiver unit is able to generate a low noise and high SNR output signal. The low noise design of the receiving unit helps the portable NMR device generate measurements with a high statistical confidence level.

[0177] In one example application, a portable NMR device can be used to measure the alpha and beta anomers in a subject's circulating blood or in a test blood sample. 1 The resonance peak generated by the CH hydrogen group protons is used to detect blood glucose concentration. In at least some embodiments, the resonance peaks of the α and β anomers are measured by using a novel method that combines water signal suppression, homonuclear decoupling, and a one-to-one mapping of the spectral peaks generated in the NMR spectrogram. In some cases, the method can also include magnetic resonance velocimetry (MRV) technology to distinguish the resonance peaks generated by the glucose molecules in the circulating blood from the resonance peaks generated by the glucose molecules in the subject's tissue.

[0178] Reference now Figure 1A , an example embodiment of an in vivo and non-invasive metabolite testing system 100A is shown. The metabolite testing system 100A includes a portable NMR device 104 for making metabolite level measurements. In particular, the portable NMR device 104 provides a convenient, compact, low-cost, and higher accuracy alternative to conventional instruments for measuring metabolite levels. As such, the portable NMR device 104 may find application in everyday consumer use (e.g., in a home setting), not just in an institutional or laboratory setting.

[0179] In at least one example application, the NMR device 104 can be used to measure important metabolite levels in diabetic patients, including glucose, glycogen, beta-hydroxybutyrate (BHB), and ketoacidosis markers. As previously described, by providing the NMR device 104 in a compact and portable form suitable for a home environment, diabetic patients and other patients with other metabolite diseases can be encouraged to use the device more frequently, thereby facilitating frequent glucose or other metabolite testing.

[0180] In other example applications, the portable NMR device 104 can be used to perform at least one of the following: drug dosage measurement, blood alcohol level measurement, opioid detection and quantification, glucose monitoring in severe burn intensive care units, metabolite level monitoring for management of metabolic diseases (e.g., PKU, IBS), etc. In various cases, the measurement of these metabolites can be performed separately or simultaneously with the measurement of glucose concentration levels.

[0181] Still reference Figure 1A , the portable NMR device 104 includes a magnet assembly having a magnet hole 106 (also referred to as a hole, cavity, bore cavity or magnet cavity) that receives a subject's finger (e.g., thumb) or a blood test sample. As shown, the hole 106 is typically located on the side of the portable NMR device 104. In at least some embodiments, the portable NMR device 104 can be designed to have a cylindrical shape and a vertical axis of symmetry to allow subjects to place their thumbs inside the hole 106 while wrapping their fingers around the outer surface of the device. To this end, the device 104 can be designed to be symmetrical to accommodate right-handed or left-handed users. In at least some cases, the portable NMR device 104 can also have a flat bottom surface 104a or a flat top surface 104b so that the device is easily located on a flat or planar surface (e.g., a household countertop). In other embodiments, the housing of the portable NMR device 104 can be non-circular.

[0182] With the subject's finger or test sample received within the magnet bore 106, the portable NMR device 104 can perform in vivo (or in vitro) and non-invasive metabolite testing in accordance with the teachings herein. The results of the test can be transmitted by the portable NMR device 104 to a remote device 112 via a network 116. The remote device 112 can be associated with a user (or subject) who is using the NMR device 104, or with a third party (e.g., a practicing physician) who is monitoring the metabolite levels of the subject or user. As non-limiting examples, the remote device 112 can be a laptop computer, a computer terminal, a mobile device, a PDA, a tablet device, or a smart phone. The network 116 can be, for example, a wireless personal area network, such as a Bluetooth® wireless network. TM A wireless local area network, such as the IEEE 802.11 family of networks, or in some cases, a wired network or communication link, such as a Universal Serial Bus (USB) interface or an IEEE 802.3 (Ethernet) network or others. In the case where the connection is a USB interface, the interface may be a USB-C interface, which supports high-speed data transfer and can also provide power to the portable NMR device 104.

[0183] In at least some cases, the portable NMR device 104 can communicate in real time or near real time with the remote device 112. In other cases, the NMR device 104 can store the collected data in a storage device for later transmission to the remote device 112.

[0184] After receiving the data from the portable NMR device 104, the remote device 112 may be configured to display the received results to the user on the display screen 114 of the remote device 112. In at least some example cases, an application may be installed on the remote device 112 that is configured to present (or display) the data received from the portable NMR device 104 to the user. The application may be, for example, an NMR spectroscopy processing software program that is configured to analyze the received data and display one or both of the raw data and the analyzed or processed data to the user. In various cases, the application may also include a graphical user interface (GUI) to display the results in a user-friendly manner.

[0185] Reference now Figure 1B , an example embodiment of an in vitro metabolite testing system 100B is shown. System 100B is similar to system 100A, but is modified for laboratory or institutional applications.

[0186] System 100B includes an NMR device 120 that replaces the portable NMR device 104. In particular, the NMR device 120 has a magnet cavity 122 facing upwards, which is surrounded by a permanent magnet assembly 126. In various embodiments, the permanent magnet assembly 126 may have a horizontal axis of symmetry. The magnet cavity 122 receives a test sample 128 for in vitro laboratory testing. The test sample 128 may be, for example, a standard 5mm diameter and 70mm NMR test tube. A mechanical sample support may be located within the cavity 122 to support the test tube in a suitable position to be measured by the NMR device 120. The NMR device 120 may also include a spectrometer 124 for receiving and processing NMR signals. In at least some cases, the spectrometer 124 may be located in a cradle base 123 of the NMR device 120. In other cases, the spectrometer may be a card inserted into a peripheral slot of a laboratory workstation, and an extended connection (e.g., an Ethernet cable) may be required to be connected to a probe located in the NMR device. In various embodiments, the stand base 123 has an upper surface having a shape complementary to the outer surface of the NMR device 120 to hold the NMR device 120 in a stable position. In at least some cases, a display screen (e.g., an LCD screen) can be located on or otherwise attached to the stand base 123. The display screen can be used, for example, to display raw data or processed data obtained by the NMR device 120, or otherwise enable a user to control the operation of the NMR device 120.

[0187] Reference now FIG. 2A to FIG. 2D , Figure 1A The general principles underlying the operation of the portable NMR device 104 will be described herein by way of example. Although the remainder of the discussion will focus on Figure 1A However, it will be appreciated that the discussion also applies to the portable NMR device 104 of Figure 1B The NMR device 120 is provided.

[0188] First reference Figure 2A , showing an exemplary 1 H Hydrogen proton (or atomic nucleus) 204. 1 The H hydrogen protons 204 may be located, for example, in a metabolite that is the subject of testing (e.g., measurement) by the portable NMR device 104. For example, 1 The H hydrogen protons may be located in glucose compounds placed in a blood sample, in the blood circulating in a subject's finger, or in the interstitial fluid of an in vitro or in vivo sample.

[0189] The portable NMR device 104 first operates by applying a uniform static magnetic field (B0) across the magnet bore 106 such that the magnetic field is oriented perpendicular to the longitudinal axis of the magnet bore 106. As described in further detail herein, the static magnetic field is generated by a combination of shims and permanent magnet assemblies located in the portable NMR device 104.

[0190] In the simplified model, the static magnetic field (B0) generated by the NMR device 104 causes 1 The H hydrogen proton ensemble is polarized and aligns the macroscopic vector magnetic moment with the direction of the static magnetic field (B0), which is usually defined along the z-axis.

[0191] The sample will contain a proton ensemble (or atomic nuclei), each of which is characterized by an independent magnetic moment. The vector sum of the magnetic moments produced by each proton in the proton ensemble is represented as the net magnetic moment (M0). Therefore, when a static magnetic field (B0) is applied, the net magnetic moment (M0) is aligned with the direction of the static magnetic field (B0). The net magnetic moment (M0) of the proton ensemble under the influence of the static field (B0) can be expressed according to equation (1).

[0192]

[0193] In an exemplary application of applying a static magnetic field (B0) to a sample containing glucose molecules, N defines the number of resonant hydrogen glucose protons in the sample (e.g., the number of alpha glucose hydrogen protons in the sample), and γ is the magnetization 1 The gyromagnetic ratio of the H proton is 2.68E+08 rad / (sec*Tesla) or 42.58 Hz / T, k is the Boltzmann constant (1.38e-23 joules / Kelvin), and is the simplified Plank constant (1.05E-34 joules per second). If N is normalized per unit volume, then M0 represents the net magnetic moment, or magnetization, per unit volume in amperes per meter or A*m 2 According to the embodiments described herein, if N is approximated as 4.2E+18 protons within the sample received in the magnet bore 106 of the portable NMR device 104, and the static magnetic field (B0) is approximately 1.68 T, then the net magnetic moment (M0) is approximated as 9.49E-08 A*m 2 / L.

[0194] Once polarization is induced by the static field (B0), the portable NMR device 104 applies a radio frequency (RF) excitation field (B1) in a direction orthogonal to the static field (B0) (e.g., in the XY plane). As further detailed herein, the RF excitation field (B1) can be generated by a set of transmission coils located within the portable NMR device 104.

[0195] In various embodiments, the RF excitation field (B1) is a circularly polarized or elliptically polarized signal applied at an angular frequency configured to excite or induce resonance in a specific type of protons (or atomic nuclei) located within the sample. The specific angular frequency at which the excitation field (B1) is applied is also referred to as "Larmor", i.e., the resonance angular frequency. Since each type of proton (or atomic nucleus) can generally be excited at a different Larmor frequency, an RF excitation field (B1) applied at one Larmor frequency can excite specific protons (or atomic nuclei) while leaving other protons (or atomic nuclei) undisturbed. In particular, as shown in equation (2), the Larmor angular frequency (ω0) is a function of both the gyromagnetic ratio of the protons and the static magnetic field strength (B0).

[0196] ω0=γB0 (2)

[0197] The Larmor frequency is also affected by a principle called "chemical shift". This is due to the local shielding of a proton from the static field (B0) by its molecular bonding environment and results in a small shift in the Larmor frequency of that proton. For example, 1 The Larmor frequency of the H protons is different from that of the protons bonded to lactic acid or water. 1 The Larmor frequency of hydrogen H. Chemical shift (δ) is reported in parts per million (ppm) and is expressed according to equation (3):

[0198]

[0199] where (f) is the resonant (or Larmor) frequency of the target proton, and f REF is the standard reference (f REF) at the resonance (or Larmor frequency) of the standard. The standard reference can be DSS (2,2-dimethyl-2-silapentane-5-sulfonate), while in other cases it can be water or other metabolites such as N-acetylaspartate (NAA).

[0200] As explained in further detail herein, the principle of chemical shift is important for the application of portable NMR device 104 in glucose concentration measurement. More specifically, chemical shift allows NMR device 104 to identify a carbon atom bonded to a known glucose carbon atom. 1 H hydrogen protons and bonds to other compounds (such as lactic acid or water) 1 H hydrogen protons. Thus, the portable NMR device 104 can distinguish the 1 The RF excitation field (B1) is applied at the Larmor frequency of H hydrogen protons.

[0201] The effect of the RF excitation field (B1) applied at the resonant frequency or Larmor frequency is to excite the protons and thereby shift the angular alignment of its magnetic moment with the axis (e.g., Z-axis) of the static magnetic field (B0) by an angular shift known as the "nutation" angle or "tilt" angle. A longer and larger amplitude RF excitation field (B1) will result in a larger misalignment of the magnetic moment (or net magnetizing moment (M0)) relative to the axis of the static magnetic field (B0) (i.e., a larger nutation angle or tilt angle). In particular, the nutation angle or tilt angle (α) generated by the RF excitation field (B1) can be expressed according to equation (4):

[0202] α=γB1τ (4)

[0203] Wherein, γ is the gyromagnetic ratio, B1 is the magnitude of the RF excitation field (B1), and τ is the duration of the excitation field (B1).

[0204] The shift in the magnetic moment also causes the magnetic moment to rotate about the axis of the static magnetic field (B0) in a motion called "precession", which occurs at the Larmor angular frequency. In general, the precession of the magnetic moment can be decomposed into two vector components: (1) a longitudinal vector component (M) along the z-axis (or the axis of the static magnetic field (B0)) and (2) a longitudinal vector component (M) along the z-axis (or the axis of the static magnetic field (B0)). z ); and (2) in the XY plane (M xy The transverse vector component in the plane perpendicular to the axis of the static magnetic field. xy ) rotates sinusoidally at the Larmor angular frequency about the axis of the static magnetic field (B0).

[0205] Reference now Figure 2B , due to the excitation field (B1), 1 The magnetic vector moment of the H proton has shifted 90 degrees into the transverse XY plane. In this position, the magnetic moment has no longitudinal component (Mz ), while the lateral component (M xy ) precesses about the z-axis at the Larmor angular frequency.

[0206] Reference now Figure 2C and 2D , once the RF excitation field (B1) is removed, the magnetization moment gradually regains equilibrium and realigns with the axis of the static magnetic field (B0) in an energy release process called "relaxation".

[0207] Similar to the decomposition of the magnetic moment (M0) during precession, relaxation is also decomposed into two components: (1) spin-lattice relaxation (also called longitudinal relaxation or T1 relaxation); (2) spin-spin relaxation (also called transverse relaxation or T2 relaxation).

[0208] Spin-lattice relaxation (T1) describes the spin-lattice relaxation of the longitudinal magnetic component (M) during realignment with the axis of the static magnetic field (B0). z ). In contrast, spin-spin relaxation (T2) describes the gradual re-growth of the transverse magnetic component (M) during realignment with the axis of the static magnetic field (B0). xy ) attenuation.

[0209] Specifically, the energy released during T2 relaxation generates a free induction delay (FID) signal or magnetic resonance signal. The FID signal is a sinusoidal signal with the Larmor frequency and a transverse magnetic component (M xy ) decay rate oscillations. For this reason, the FID signal is usually characterized by a decaying exponential envelope that decays at a T2 time constant. In various cases, T2 relaxation can also include T2* relaxation. T2* represents the actual (or effective) relaxation caused by external factors such as inhomogeneities of the static magnetic field (B0). T2* relaxation shortens (or suppresses) the free induction decay signal.

[0210] Now briefly refer to Figure 2E , shows a largely simplified model of T1 and T2 relaxation. As shown, after the proton is excited at a 90-degree angle, the net magnetic moment starts out entirely in the transverse plane. As the magnetic moment returns to equilibrium, the vector component along the z-axis (M z ) gradually increases (i.e., T1 relaxation), while the vector component in the XY plane (M xy ) gradually decays to zero (i.e., T2 relaxation). The longitudinal component (M z ) and the lateral component (M x ) and (M y ) can be expressed according to equations (5), (6) and (7):

[0211]

[0212]

[0213]

[0214] Where Ω = ω0-ω is the angular rotation offset frequency, T1 is the longitudinal component (M z ) growth time constant, T2 is the transverse component (M xy ) decay time constant.

[0215] As indicated by at least equations (6) and (7), the M of the FID signal x (t) and M y (t) components are 90 degrees out of phase. Therefore, M x (t) and M y (t) of the combined vector components (e.g., M xy ) is interpreted as a circularly polarized FID signal.

[0216] Now briefly refer to Figure 2F , shows an example graph 200F showing an example FID signal generated as a result of T2 relaxation. As shown, the FID signal is represented by a decaying envelope at a T2 time constant.

[0217] In the embodiments described herein, metabolite levels are generally measured by measuring the FID signal as an induced AC voltage on a set of receiver coils located within the magnet bore 106 of the portable NMR device 104. In particular, the voltage amplitude S of the magnetic resonance or FID signal is emf It can be expressed according to equation (8):

[0218]

[0219] Where n is the number of turns of the receiving coil solenoid, r is the radius of the turns of the receiving coil solenoid, l is the half length of the receiving coil, M o is the spin angular moment per unit volume or the magnetization in amperes per meter (calculated above), V s is the sample volume, and μ0 is a constant of the permeability of free space (1.26E-06T*m / A=H / m). In at least some embodiments described herein, the portable NMR device 104 is configured to generate a static magnetic field of 1.68T, receiving approximately 3.5×10 -6 m 3 The sample volume (V s ), and using a receiver coil with 10 turns and a radius of 0.012 meters, the half length of the coil is 0.0150 meters. Under these parameters, the received voltage amplitude of the resonant signal will have a peak amplitude of approximately 50nV.

[0220] Equation (8) also shows that the magnetic resonance signal S emf The voltage is proportional to the square of the static magnetic field (B0). This is because the Larmor angular frequency (ω0) and the magnetic moment (M0), both included in equation (8), are proportionally related to the static magnetic field (B0) (e.g., see equations (1) and (2)). Therefore, a stronger static magnetic field (B0) will produce a larger amplitude magnetic resonance signal S emf , and by extension, produces a signal with a higher signal-to-noise ratio (SNR). According to the teachings provided herein, the ability of the portable NMR device 104 to produce a static magnetic field greater than 1.5 T facilitates the device's ability to produce highly accurate measurements with low SNRs, and facilitates the ability to use only one or two scans of a subject's finger or test sample.

[0221] Reference now Figure 3A , shows an NMR spectrum 300A, which shows the various protons located in (or bonded to) the glucose molecule after taking into account the chemical shift characteristics of each proton. 1 The Larmor frequency of H hydrogen protons can be measured by measuring the FID signal (which consists of individual 1 The FID signal is then converted to the frequency domain to produce the NMR spectrum 300A.

[0222] As shown in the figure, the glucose molecule has twelve 1 H hydrogen protons, each 1 The H protons have different bonding environments. 1 Of the H protons, only seven are directly visible to NMR (i.e., the RF emission frequency in the FID signal is observable during relaxation), because the other five protons are in hydroxyl groups, so-called "water exchangeable" and resonate at the same frequency as the hydrogen protons in water molecules. 1 Each of the H hydrogen protons resonates at a different frequency. At 302, a component was observed at 5.22 ppm, the α-glucose (α-Glc) anomer 1 The CH hydrogen group protons are approximately 0.5 ppm different from the resonance of water and other metabolites. The resonance peak 304 at 4.72 ppm corresponds to the hydrogen protons located in water, while the reference signal at 0 ppm corresponds to DDS (2,2-dimethyl-2-silapentane-5-sulfonate). The peak 306 at 4.634 ppm corresponds to the β-glucose (β-Glc) anomer 1 The CH hydrogen group protons resonate very close to the water signal peak. In the region 308 between 3 ppm and 4 ppm, the remaining glucose 1 H hydrogen protons resonate overlappingly with other metabolite protons.

[0223] Although NMR spectrum 300A shows a one-to-one mapping of spectral lines to glucose hydrogen protons, in reality, a one-to-one mapping may not be possible due to a phenomenon known as "homonuclear spin coupling" or "spin-spin coupling." Spin-spin coupling causes a single spectral line associated with a single glucose hydrogen proton to be resolved into two or more spectral lines. More specifically, spin-spin coupling is caused by magnetic interference generated by nearby or adjacent hydrogen proton spins in a glucose molecule. As described in further detail herein, mitigating the effects of spin-spin coupling requires the application of spin-spin decoupling techniques.

[0224] In at least some embodiments, the portable NMR device 104 can determine the blood glucose concentration by applying a one-to-one mapping of the seven visible hydrogen protons located in the glucose molecule to the corresponding resonant signal peaks in the NMR spectrum. The method applies a broadband spin-spin decoupling technique, which resolves multiple spectral lines associated with one hydrogen proton into a single spectral line. More specifically, in order to resolve the glucose spectral lines between 3ppm and 4ppm (region 308) from other metabolites, the portable NMR device 104 is configured for high spectral resolution and applies a static magnetic field with a uniformity of 0.01ppm.

[0225] In other embodiments described in further detail herein, the portable NMR device 104 may also use only the α and β anomers. 1 The one-to-one mapping of CH hydrogen group protons to the corresponding resonance signal peaks generated by these protons in the NMR spectrogram determines the blood glucose concentration. This method avoids the complexity of many closely approaching spectral lines due to spin-spin coupling in the decomposition region 308. The portable NMR device 104 measures the resonance signal amplitude of the α and β anomers by applying a spin-spin decoupling method, which resolves the doublet spectral lines (i.e., two peak spectral lines) of the α and β anomers into a single spectral line. Since the β anomer is very close to the water signal, a water suppression method is used before exciting the α and β anomer hydrogen protons to reduce the interference of the water signal in the measurement of the α and β anomers in the spectral domain.

[0226] Measuring the resonance peaks of the α and β anomer hydrogen protons enables the NMR device 104 to generate measurements with a high statistical confidence level of 0.95 and an accuracy (ε) of + / - 2% of the true level of blood glucose concentration (assuming a nominal average blood glucose concentration of 5.5 mM (millimoles per liter)). The statistical confidence level of 0.95 at an accuracy (ε) of + / - 2% allows accurate insulin dosing decisions to be made and represents an improvement over conventional fingerstick blood glucose meters, which achieve an accuracy of 5% to 20% at a confidence level of 0.95. In particular, the portable NMR device is able to generate measurements with this high level of accuracy using only one or two scans of the subject's finger or test sample by the NMR device 104, and in a competitive time frame of no more than 12 seconds. In various embodiments, higher measurement accuracy and / or statistical confidence levels may be achieved when the blood glucose concentration is greater than 5.5 mM.

[0227] According to the teachings herein, a statistical confidence level (CL) of 0.95 is determined based on the error function expressed in equation (9):

[0228]

[0229] where Δf is the measurement bandwidth for metabolite quantification, σ n is the noise in the received signal, ε is the required accuracy of the system (e.g. 0.02), S emf is the voltage amplitude of the received magnetic resonance signal (eg 60 nV).

[0230] The noise (σ) in the received signal in equation (9) n ) comes primarily from the receiver electronics, which is the primary source of noise. As explained in further detail herein, to achieve a confidence level of 0.95, the portable NMR device 104 uses receiver electronics configured to have a 1.1 The noise level (σ) of the reference input (RTI) n ). It is assumed that the noise in the received electronic equipment is spectrally uniform and has a Gaussian-like amplitude distribution.

[0231] The measurement bandwidth Δf in equation (9) is a function of the FID (or T2 relaxation) signal. An approximate value of Δf based on the transient decay of the transverse T2 magnetization is expressed according to equation (10):

[0232]

[0233] where ΔB0 represents the root mean square average of the static magnetic field B0 spatial inhomogeneity across the sample or subject's finger. To obtain a statistical confidence level of 0.95, the measurement bandwidth was minimized using the homonuclear decoupling method, which reduces the doublet spectral line to a single spectral line occupying the smallest frequency band.

[0234] The confidence level is also related to the magnetic resonance signal (S emf ) is positively correlated with the voltage of the static magnetic field (B0), which, as previously described, is proportional to the square of the static magnetic field (B0). Therefore, increasing the magnitude of the static magnetic field (B0) generates measurements with a higher statistical confidence level. In various embodiments described herein, a static magnetic field (B0) between 1.5T and 2T is used to generate measurements with a 0.95 confidence level.

[0235] It will be appreciated that measuring only the resonance peak of the alpha anomer is insufficient to achieve a statistical confidence level of 0.95 with an error of + / - 2%.

[0236] Reference now Figure 3B , showing the α-anomer-only 1 Graph 300B of statistical confidence levels for glucose concentration measurements of H hydrogen proton resonance. In particular, graph 300B shows confidence levels as a function of static magnetic field strength and measurement noise (as receiver reference input (RTI)). Graph 300B assumes a nominal blood glucose concentration level of 5.5 mM.

[0237] As shown, with an error of + / - 2%, the static magnetic field (B0) strength is set to 1.68T (e.g., below 2T to comply with various consumer safety regulations) and the receiver noise is at In the case of low values ​​of , only a confidence level of 0.65 can be obtained. Even when the static magnetic field (B0) strength is increased to a maximum of 2T, the confidence level does not reach a statistical confidence level of 0.95 (i.e., the confidence level is just over 0.8). In addition, and as expected, the confidence level decreases with higher receiver noise. For example, with a receiver noise of 1.1 nV / √Hz (RTI) and the static magnetic field (B0) strength set to the maximum value of 2T, the highest confidence level achieved is just under 0.8.

[0238] Therefore, unless the α anomer is obtained 1 It is not feasible to achieve a confidence level of 0.95 with an error of ±2% using only the α anomer, since multiple measurements of the H resonances are averaged. 1 Averaging N measurements of the H resonance can improve the measurement accuracy and signal-to-noise ratio (SNR) by a factor of √N. 1H measurements take 5 to 7 seconds (i.e., the α anomer 1 H T1 relaxation time), so it is not possible to use only the α anomer 1 One or two measurements of the H resonance are used to achieve the desired error rate and confidence level within the competitive time frame of 12 seconds, which is the time required for current finger-prick glucose tests to complete a measurement.

[0239] In addition, by rotating the α anomer at the appropriate Ernst angle 1 The net magnetic moment of H realizes the α anomer 1 Multiple consecutive measurements of the H resonance signal to ensure the best SNR performance. Ernst angle (α E ) is usually expressed by equation (11):

[0240]

[0241] Among them, T R is the scan repetition time, T1 is the α anomer 1 The longitudinal relaxation time of H. During the scan repetition time T R When T1 is equal to (i.e., in order to increase the number of measurements of the resonance signal within a given time interval), the optimal Ernst angle is about 68.4°, which improves the measurement SNR by up to 10 dB (relative to T R ≥5T1 and rotated at 90° nutation angle). However, the α anomer is generated by rotating the net magnetic moment at the Ernest angle 1 Multiple consecutive measurements of the H resonance signal can also reduce the quantitative accuracy because of the difference between the measured free induction decay amplitude and the α anomer. 1 A dependence is introduced between the T1 relaxation times of H protons.

[0242] Reference now Figure 3C , showing the α and β anomers 1 Graph 300C of the statistical confidence level of glucose concentration measurements of H hydrogen proton resonance. In particular, curve 300C shows the confidence level as a function of static magnetic field strength and measurement noise (as receiver reference input (RTI)). Curve 300C also assumes a nominal blood glucose concentration level of 5.5 mM.

[0243] As shown in the figure, when the receiver noise is lower than And a confidence level of 0.95 with an error of + / -2% can be achieved when the static magnetic field (B0) strength is set to at least 1.68 T. Thus, measuring both alpha and beta resonances allows for highly accurate measurements with high statistical confidence levels, in contrast to measurements that rely solely on alpha proton resonance. In particular, these measurements can be obtained within a competitive time frame and using only one or two scans of a subject's finger or test blood sample.

[0244] Reference now Fig.10 , shown for α- and β-glucose anomers 1 An exemplary process flow 1000 of a method for measuring blood glucose concentration based on a resonance peak of H resonance.

[0245] In act 1002, a static magnetic field (B0) is applied to a subject's finger or a test sample to induce polarization of an ensemble of atomic nuclei located therein.

[0246] In actions 1004 and 1006, a method for suppressing water signals is applied. In particular, due to the β anomer 1 The CH hydrogen proton resonance (δ = 4.634 ppm) is close to the spectral modulation sideband generated by the water signal resonance peak (δ = 4.72 ppm), so the excitation of the β-anomer hydrogen proton also has the risk of exciting the water hydrogen proton. This may cause the FID signal generated by the water signal to interfere with the measurement of the FID signal generated by the β-anomer hydrogen proton.

[0247] Thus, in act 1004, the water suppression method includes first applying an RF excitation pulse (B1) to the subject's finger or test sample to rotate the magnetization of hydrogen protons located in the water to the transverse plane. This is similar to the microscopic Figure 2B In various embodiments, the magnetization of the water compound is rotated into the transverse plane by rotating the water compound at the water resonant frequency f H2O This is achieved by applying a long frequency selective pulse (B1) of sufficient duration and amplitude.

[0248] In at least some embodiments, the long frequency selective pulse (B1) is gated with a rectangular pulse and is at the water resonance frequency f H2O The RF excitation pulse has a frequency domain excitation profile expressed by equation (12):

[0249] |sinc(πΔfτ)| (12)

[0250] where Δf is the offset frequency from the center frequency of the water resonance, and τ is the duration of the rectangular pulse calculated according to equation (13):

[0251] τ=(f H2O -fβ-glc ) -1 (13)

[0252] where f H2O is the resonant frequency of water protons, f β-glc is the resonant frequency of the β anomer. Therefore, by centering the pulse on the water resonant frequency and defining the pulse duration (τ) relative to equation (13), the β anomer should be undisturbed (or not excited) by the pulse.

[0253] Now briefly refer to Figure 3D , shows the transverse magnetization (M0) normalized to the original longitudinal magnetization (M0) after applying a frequency selective pulse (B1). xy ) is a graph 300D of the desired amplitude frequency response.

[0254] As shown in the figure, the |sinc(πΔfτ)| pulse (shown as the solid curve) excites the water signal, while the β-glucose anomer 1 The chemical shift of the CH group hydrogen proton is located at the zero of the sine wave and is not disturbed. α anomer 1 Although the chemical shift of the CH group hydrogen proton is close to the zero point of the sine wave, it undergoes about 4% resonant excitation due to its position on the sine wave lobe.

[0255] To ensure that both the α and β anomer hydrogen protons are at frequency zero, in some embodiments, the RF excitation pulse (B1) may be a triangular pulse of duration (2τ) gated at the water resonance frequency f H2O The RF excitation pulse has a frequency domain excitation profile expressed by equation (14):

[0256] |sinc 2 (|πΔfτ)| (14)

[0257] where Δf is the offset frequency from the center frequency of the water resonance, and the pulse duration τ is defined according to equation (13).

[0258] Still reference Figure 3D , the triangular excitation generated by equation (14) is shown by the dashed line. As shown, both the α and β anomer hydrogen proton resonance peaks are now located at the frequency zero. Therefore, the triangular pulse response ensures that the metabolite magnetization will not be disturbed even if there are small errors in the zero frequency location.

[0259] The amplitude of the long frequency selective pulse (B1) that excites the rotation of the water signal into the transverse plane is expressed according to equation (15):

[0260]

[0261] Where τ is the pulse duration and γ is the gyromagnetic ratio of the target nucleus. For a pulse duration (τ) of 163 ms (e.g., determined according to equation (13)), the required RF excitation field (B1) is about 40 nanovolts at the gyromagnetic ratio of hydrogen. However, a single pulse with an amplitude of 40 nanovolts may be too small to be practically achieved.

[0262] Therefore, in various embodiments, the RF excitation signal (B1) may be segmented into a series of high energy short pulses that approximate the amplitude expressed in equation (15). For example, the excitation signal may be implemented using a Delay Variable Precession Tailored Excitation (DANTE) technique.

[0263] The DANTE technique splits a single rectangular pulse (which achieves a 90° shift in the transverse plane) into a series of p ), where successive pulses are separated by a duration (Δt). Each pulse in the train of n pulses achieves a 90° / n rotation of the water signal. Assume Δt>>t p , then the amplitude of n short pulses increases by Δt / t relative to the original single pulse p times. The duration (τ) of the DANTE pulse sequence can be expressed according to equation (16):

[0264] τ=nt p +(n-1)Δt (16)

[0265] The DANTE sequence leads to p +Δt) -1 Hz apart. Thus, Δt is adjusted to ensure that the excitation frequency of the DANTE pulse does not eventually overlap with the excitation pulses of the α or β anomers. In various embodiments, extending the DANTE technique to triangular excitation pulses (as described above) requires the application of envelope modulation.

[0266] Now refer back to Fig.10 , a dephased pulsed field gradient (PFG) is applied to the water signal at act 1006. In particular, the PFG attenuates the FID signal generated by the water signal and allows the resonances of the α and β anomers to be measured without interference from the water signal.

[0267] Now briefly refer to Figure 3E , shows an example illustration of the effect of pulsed field gradient (PFG) on a water signal. As shown, after applying a 90° pulse to the water signal, the net magnetic moment of the water signal rotates into the transverse plane. The spins are coherent after the 90° pulse and, therefore, precess around the z-axis at the same rate to produce a net magnetic moment.

[0268] PFG is a gradient magnetic field applied as a static magnetic field (B0) along the axis and has a variable gradient field strength in the XY plane. More specifically, PFG exposes different spins to different fields according to their spatial positions and causes the spins to lose coherence (e.g., spins precess around the z-axis at different rates). Therefore, the average sum of the individual spins no longer coherently adds to produce a strong net magnetic moment in the transverse plane. In this way, the net magnetic moment of the water signal "weakens", and the resonance signal generated by the water molecules is "attenuated" or "suppressed". The magnetic moments of the α and β end isomer hydrogen protons can now be excited separately without being disturbed by the water resonance signal. In various embodiments, the combination of action 1004 and action 1006 results in an attenuation of the water signal between 12 dB and 20 dB.

[0269] As previously described, relaxation back to the z-axis occurs at a T1 exponential rate (see equation (5)). The T1 relaxation of water molecules in blood is about 1.4 seconds. Therefore, PFG should have a pulse duration significantly less than 1.4 seconds. In at least some embodiments, water suppression can also be achieved by using hyperbolic secant pulses (e.g., adiabatic RF pulses). However, hyperbolic secant pulses produce magnetization excitation over a limited frequency band, so their time domain pulses are infinite and must be truncated, a process that increases the transition band width and introduces frequency domain variations in the magnetization. This is in contrast to the use of the above-mentioned techniques, which only limit the frequency domain magnetization profile at the point of interest and result in a truncated limited time range RF pulse that eliminates the truncation.

[0270] In action 1008, a second RF excitation pulse (B1) is applied to rotate the α and β anomers 90° into the transverse plane. In at least some embodiments, the second RF excitation pulse is applied at a maximum RF amplitude and a minimum duration. This allows for mitigation of transverse relaxation decay at an exponential rate T2* that occurs prior to free induction decay (FID) measurements. In at least some embodiments, the duration of the applied RF excitation pulse is less than 1.5 milliseconds.

[0271] At act 1010 , lower power continuous wave (CW) radiation is applied to achieve homonuclear decoupling. Act 1010 may be performed at least partially concomitantly with act 1008 .

[0272] CW radiation separates the α and β anomers 1 The doublet of H protons is resolved into singlets. For example, CW radiation resolves the α anomer 1The two spectral lines of the H protons, centered at the 5.223 ppm chemical shift point and separated by 3.8 Hz, are resolved into a single spectral line at the chemical shift point. As previously described, homonuclear decoupling allows a one-to-one mapping of the spectral lines in the received FID signal. As previously described, homonuclear decoupling also minimizes the required frequency measurement bandwidth and, in turn, increases the statistical confidence level of the measurements obtained by the portable NMR device.

[0273] because 1 α and β anomers in CH groups and glucose 2 The α and β anomers of the CH group are coupled, thus forming two anomers 2 The CW radiation was applied at the resonance frequencies of the hydrogen protons of the CH groups (δ = 3.519 ppm and δ = 3.23 ppm) to achieve highly selective decoupling.

[0274] In at least some embodiments, the CW radiation is applied at a power of less than 12 mW / gram of tissue.

[0275] At action 1012, and in at least some embodiments, magnetic resonance velocimetry (MRV) techniques are applied to distinguish between resonance signals generated by glucose molecules located in the plasma and otherwise affected by arterial or venous blood flow, and resonance signals generated from glucose molecules that may be located, for example, in human tissue and are in a fixed state. In particular, MRV may find particular application in which blood glucose concentration levels are measured in a subject's finger (i.e., rather than a test blood sample) received within the aperture 106 of the NMR device 104. In these cases, MRV ensures that the measured blood glucose concentration level is limited to the glucose concentration in the blood circulating in the subject's finger, and does not additionally include the glucose concentration level of glucose located in the subject's tissue.

[0276] More specifically, MRV is used to distinguish between stationary proton spins and spins that are part of the fluid flow. Since arterial blood flows at velocities between 4.9 cm / s and 19 cm / s, and venous blood travels at velocities between 1.5 cm / s and 7.1 cm / s, MRV can be used to distinguish between glucose that is affected by arterial or venous blood flow and glucose that is otherwise stationary (i.e., located in tissues).

[0277] Now briefly refer to Figure 3F and 3G , the basic principles of MRV will be explained in further detail here.

[0278] Now first refer to Figure 3F, an exemplary bipolar gradient pulse (BGP) is shown, which can be applied to a subject's finger as part of an MRV technique (also known as a flow encoding gradient). The BGP generally functions in a similar manner to the pulsed field gradient applied in action 1006. More specifically, the BGP is a gradient magnetic field applied along the axis of the static magnetic field (B0) and has a variable gradient field along the x-axis. The gradient varies along the x-axis because the blood circulation within the finger placed in the bore 106 of the NMR device 104 will flow primarily along the x-axis (i.e., defined relative to the NMR device 104).

[0279] like Figure 3F As shown, the BGP includes a first positive component 302F (ie, having a positive bipolar gradient) followed by a second negative component 304F (ie, having an inverted negative bipolar gradient).

[0280] Reference now Figure 3G , shows the effect when BGP is applied to a fixed proton. In particular, after the first positive component 302F of BGP is applied, the precession rate of the transverse component of the proton's magnetic moment is perturbed by the gradient field (e.g., the precession velocity may be accelerated as a result of BGP, resulting in a phase shift of the transverse component). In particular, the phase shift induced in a group of proton spins by BGP at a given time (τ) may be expressed according to equation (17).

[0281]

[0282] in is the phase shift at time (τ), is the initial phase, and ω(t) is the instantaneous Larmor frequency of the population of proton spins, determined according to equation (18).

[0283] ω(t)=γ(B0+G x (t)x(t)) (18)

[0284] Where B0 is the strength of the static magnetic field, and G x (t)x(t) is the intensity of the BGP along the z-axis and is a function of the position along the x-axis. In general, phase accumulation due to the constant static field (B0) term can be eliminated by demodulation at the spectrometer receiving unit of the NMR device 104.

[0285] exist Figure 3G In the example shown, the BGP causes the transverse component to experience a phase shift of approximately 5 radians at 302G. After applying the first positive component 302F, the second negative component 304F of the BGP is applied. The second negative component has the effect of reversing (or inverting) the disturbance caused by the first positive component 302F to the precession. Thus, as in Figure 3GAs observed in Figure 2, during 304 G, the 5 mrad phase shift of the transverse component of the magnetic moment is reversed, resulting in a net phase shift of zero. Thus, the magnitude and duration of the positive and negative components of the BGP are designed to advance and retard the fixed spin resonance phase by exactly the same amount to return the spin to its original phase.

[0286] like Figure 3G The effect of applying BGP to protons moving at a constant velocity (e.g., α and β anomer glucose hydrogen protons that may flow through the veins and arteries of a subject) is also shown. Figure 3G Assume that 0.2 mT / cm is applied along the x-axis (G x ) is applied to the protons traveling at a constant velocity of 5 cm / sec.

[0287] like Figure 3G As shown, when the positive component 302F of BGP is applied, the magnitude of the phase shift accumulated in the transverse component of the magnetic moment becomes quadratic (e.g., 306G). This is because as the proton moves along the x-axis (i.e., through an artery or vein), the proton is changing its spatial position relative to the gradient field (i.e., changing along the x-axis). Therefore, during the pulse duration of the positive component of BGP, the moving proton is constantly affected by the variable magnetic field. When the negative component 304F is subsequently applied, the accumulated phase shift to the magnetic moment as a result of the positive component does not reverse (i.e., as was the case when the proton was fixed). This is because the proton is now displaced along the gradient axis and is no longer experiencing the exact inverse gradient that was applied during the positive component 302F. Therefore, during Figure 3G In FIG. 1 , the inverse component produces an inverse or net phase shift of approximately -3 millirads (eg, 308G).

[0288] In view of the foregoing, protons that experience a constant velocity (e.g., protons in circulating blood) experience a net phase shift after application of BGP, while protons that are stationary (e.g., located in tissue) experience zero net phase shift after BGP. Thus, this property is used to distinguish between resonance signals generated by glucose molecules located in the blood and resonance signals generated by glucose molecules located, for example, in tissue.

[0289] In various embodiments described herein, where MRV is desired, method 1000 is applied twice to resolve blood glucose from tissue glucose. In particular, different BGPs are applied between the first and second iterations of method 1000. Typically, an inverted version of the original BGP is applied in the second iteration. The results (e.g., resonance signals) generated from the two iterations of method 1000 can be subtracted to provide a measurement specific to streaming spin. Alternatively, the results of the two iterations can be added to provide a measurement of glucose concentration in finger tissue.

[0290] At act 1014 , the FID signals of the alpha and beta hydrogen anomers are measured based on T2 relaxation. In particular, and as explained in further detail herein, the FID signals are detected by a set of receiver coils located within the magnet bore 106 of the portable NMR device 104 .

[0291] At act 1016, the measured FID signal is converted to the frequency domain to generate an NMR spectrum (eg, Figure 3A ). In at least some embodiments, the frequency domain conversion is achieved by discrete Fourier transform (DFT). DFT can be facilitated by resolving the resonance peaks of α and β anomers into individual spectral peaks. In other embodiments, as non-limiting examples, the frequency domain conversion can also be achieved by discrete cosine transform (DCT) or discrete sine transform (DST).

[0292] At act 1018, the alpha and beta anomer resonance amplitudes determined from the NMR spectrum are converted to glucose concentrations.

[0293] In at least some embodiments, glucose concentration conversion is achieved by correlating (a) the resonance peaks of the α and β anomers (e.g., determined from an NMR frequency plot) to (b) known reference glucose concentration levels. For example, a memory unit located in the portable NMR device 104 can store correlation information between specific resonance amplitudes and known reference glucose concentration levels. Because the resonance amplitude is linearly related to the glucose concentration level, interpolation can be used in the absence of direct correlation information in the memory unit.

[0294] In other embodiments, the conversion is performed in real time using the ERETIC (Electronic Reference for In Vivo Concentration) method. ERITEC generates artificial electronic reference resonance peaks in the NMR frequency map, which are then used to determine absolute concentration levels.

[0295] In other embodiments, the glucose concentration level can be determined without direct reference. For example, glucose can be measured 1 The ratio of the α and β resonances of the CH groups. In human blood, the nominal anomer resolution is 36% α to 64% β. Therefore, the volume V can be determined from equation (19) s The number of hydrogen protons of the α-anomer glucose in the sample:

[0296] N α-glc =0.36N A c glc V s (19)

[0297] Where N A is Avogadro's number, c glcis the average blood glucose concentration, measured in moles per liter, with a normal physiological value of 5.5x10-3mol / L (or 99mg / dL), while V s is the volume of the sample received in the portable NMR device. For example, assuming V s 2.00E-06m 3 , then the number of hydrogen protons of the α-anomer glucose in the sample is N α-glc , which is calculated to be 2.38E+18 protons.

[0298] The anomeric ratio varies depending on the glucose concentration, temperature, and pH of the circulating blood or blood test sample. In various embodiments, temperature and pH can be measured so that the glucose concentration can be determined from the anomeric ratio measured in the free induction decay signal. For example, intracellular pH can be measured in vivo by stimulating carnosine molecules in a subject's finger or test blood sample and measuring the resulting FID signal (e.g., see RAdGraaf [2]).

[0299] In each case, measuring both the α and β anomers (rather than just one of the anomers) helps to improve the repeatability of the measurement and reduce errors due to variations in the anomeric ratio.

[0300] Still in other embodiments, spectral editing can be used to distinguish glucose from macromolecules such as glycosylated proteins. This function is important because glycosylated proteins typically have NMR resonances that substantially overlap with the chemical shift of glucose. Diabetic disease progression is characterized by an increase in glycosylated proteins over time. Typically, metabolites have much higher T1 values ​​than macromolecules that undergo faster spin-lattice relaxation, and this difference can be used to null the macromolecule signal. A non-selective inversion recovery sequence (180°-td) before measuring metabolites can be used to null the macromolecule response. In various embodiments, the recovery delay is typically set to td=T1_mm*In(2) so that the net magnetization of the macromolecule is zero when the metabolite is measured.

[0301] Now briefly refer to Figure 3H , schematically showing the α- and β-glucose anomers 1 Method for measuring blood glucose concentration by CH hydrogen group proton resonance. As shown, a first 90° pulse is applied to rotate the magnetization of water hydrogen protons to the transverse plane. Then, a dephasing gradient is followed after the first pulse to attenuate the water signal. A second 90° pulse is then applied to rotate the α and β anomers of glucose. 1 The magnetization of the CH hydrogen protons rotates into the transverse plane. At least partially simultaneously with the second 90° pulse, the α and β anomers of glucose 2CW radiation is applied at the resonance frequency of the hydrogen protons of the CH group to achieve homonuclear decoupling. At least in some cases, a bipolar gradient field (BGF) (or flow encoding signal) is applied to the subject's finger to distinguish the glucose hydrogen resonance signal generated by the glucose in the circulating blood from the resonance signal generated by the hydrogen protons of the fixed glucose molecules (e.g., located in the tissue). Finally, the FID signal is measured. In the case of using the MRV technique, the sequence can be repeated a second time with an inverted BFG to properly resolve the glucose hydrogen resonance signal generated by the glucose in the circulating blood.

[0302] Reference now Figure 4A , showing another aspect of the teachings of this invention Figure 1A A simplified block diagram 400A of an example embodiment of a portable NMR device 104 is shown. As shown, the portable NMR device can be implemented to generally include: a processor unit 404 in communication with a communication module 432, a power module 412, a memory unit 414, a transmitting unit 416, and a receiving unit 420, and a shimming control unit 422. In at least some embodiments, the processor unit 404 can further communicate with one or more of a touch sensor 426, a tap sensor 428, and a servo motor 430.

[0303] According to the teachings provided herein, the processor unit 404 can be configured to execute multiple instructions to control and operate the various components of the portable NMR device 104. In some embodiments, the communication module 432 can be used to send instructions from the remote device 112 to the processor unit 404. In other embodiments, the processor unit 404 can be preconfigured with specific instructions. The preconfigured instructions can be executed in response to a specific event or a specific sequence of events or at specific time intervals. The processor unit 404 can also be configured to receive data from the various components of the NMR device 104 and use the data to make specific determinations, as described in further detail herein. The determination results can then be stored in the memory unit 414 and / or the determination results can be sent to the communication module 432 for transmission to the remote device 112.

[0304] Memory unit 414 may be, for example, a nonvolatile read-write memory that stores computer-executable instructions and data, as well as a volatile memory (eg, random access memory) that may be used as working memory by processor unit 404 .

[0305] The power module 412 may be, for example, a battery capable of supplying power to the portable NMR device 104 for a predetermined period of time. For example, the battery may be a compact battery, such as a lithium-ion (Li-Ion) battery, configured to be housed within the portable NMR device 104. In some other embodiments, the power module 412 may be an inductive power module that may receive wirelessly transmitted power and provide power to the portable NMR device 104. In other embodiments, the power module 412 may be recharged using a USB port located on the portable NMR device 104.

[0306] The communication module 432 may be configured to send data or information to the remote device 112, or to receive data or information from the remote device 112. The communication module 432 may, for example, include a wireless transmitter or transceiver and an antenna. In some embodiments, the communication module 432 may receive instructions or data from the remote device 112 and send the instructions or data to the processor unit 404. Thus, the communication module 432 may be configured to provide duplex communication.

[0307] The transmit unit 416 (also referred to as the spectrometer transmit unit 416) can be coupled to the transmit coil 408a of the probe device 408. The transmit coil 408a can be located within the magnet bore 402, which is similar to the magnet bore 106 of FIG. 1. In various embodiments described in further detail herein, the transmit unit 416 can be configured to drive the transmit coil 408a to generate a pulsed RF excitation field (B1) that induces resonance within a target proton ensemble in a subject's finger or in a sample received in the magnet bore 402. The frequency or intensity of the RF excitation field (B1) can be determined, for example, by the processor unit 404 according to the previously described techniques.

[0308] The receiving unit 420 (also referred to as the spectrometer receiving unit 420) can be similarly coupled to the receiving coil 408b of the probe device 408. The receiving coil 408b can also be located within the bore 402. In various embodiments described in further detail herein, the receiving coil 408b can be configured to receive a signal (e.g., a FID signal) generated by a sample located within the bore 402. The receiving coil 408b can then send the signal to the receiving unit 420, which reduces signal noise, and then send the signal to the processor unit 404 for further analysis. In other embodiments, the processor unit 404 can pass the received signal to the communication module 432, which can then send the signal to the remote device 112 via the network 106. The application installed on the remote device 112 can be configured to analyze or process the raw signal data to, for example, determine a metabolite level.

[0309] Since NMR spectroscopy requires a uniform static field to be generated across the magnet bore 402, a shimming control unit 422 may be coupled to the processor unit 404. The shimming control unit 422 is in turn coupled to a set of shimming coils 424 disposed within the magnet bore 402. As explained in further detail herein, the shimming control unit 422 drives the shimming coils 424 to generate a compensation magnetic field within the bore 402 to compensate for the spatial inhomogeneity of the static magnetic field (BO) generated by the permanent magnets located in the portable NMR device 104. As previously described, the shimming coils 424 may also be used to generate pulsed field gradients (PFG) and bipolar pulsed gradients (BPG) used during water signal suppression and MRV, respectively.

[0310] In at least some embodiments, the portable NMR device 104 may also include one or more of a touch sensor 426 , a tap sensor 428 , and a servo motor 430 .

[0311] A touch sensor 426 (also referred to as a proximity sensor 426) may be located inside the well 402 and may be used to detect the presence of a subject's finger or test sample contained within the well 402. In at least some embodiments, the touch sensor or proximity sensor 426 may also be used to detect whether the finger or test sample is correctly positioned within the well 402. If it is determined that the finger or sample is correctly positioned, the sensor 426 may send a signal to the processor unit 404 to instruct the processor unit 404 to activate various hardware components of the portable device 104 to perform a metabolite test. In at least some cases, more than one touch sensor 426 may be included to more accurately determine whether the subject's finger or test sample is precisely positioned within the well. In various cases, the touch sensor 426 may be a capacitive device that provides an analog value reading. Therefore, if the touch sensor 426 detects that the subject's finger or test sample is correctly positioned, the touch sensor 426 may produce a higher analog reading than if the subject's finger or touch sensor is incorrectly positioned.

[0312] The touch sensor 426 may also include a fingerprint reader that identifies the tag in the in vivo measurement. For example, the touch sensor 426 may be located within the magnet hole 402 and may receive a subject's finger located within the hole 402. When data or measurements are collected by the portable NMR device 104, the touch sensor 426 may tag the data with information about the subject's fingerprint. In at least some embodiments, the information may then be stored in, for example, a memory unit 414 for later retrieval. In the case where multiple subjects use the portable NMR device 104, it may be useful to tag the metabolite test results with an identity tag (i.e., data may be stored and classified according to each user's respective fingerprint information). In other embodiments, the portable NMR device 104 may be configured to associate the user's fingerprint with a corresponding remote device 112. The portable NMR device 104 may accordingly use information from the fingerprint reading to determine which remote device 112 should receive the metabolite test results.

[0313] A knock sensor 428 may be incorporated into the portable NMR device 104 to provide a convenient mechanism for powering up the device. For example, a subject may touch the knock sensor 428, and in turn, the knock sensor 428 may activate the processor unit 404. In at least some embodiments, the knock sensor 428 may be an accelerometer.

[0314] In at least some embodiments, the magnet bore 402 may also include a movable or sliding door at the opening of the magnet bore 402. When testing is to be performed, the sliding door opens to provide access to the bore 402. Thus, activation of the knock sensor 432 may result in activation of the servo motor 430, which opens the door to the bore 402.

[0315] Reference now Figure 4B , a circuit diagram 400B is shown which illustrates the various components of the block diagram 400A in more detail.

[0316] As shown, the processor unit 404 may include a microcontroller 404a, a single board computer (SBC) 404b, and a field programmable gate array (FPGA) 404c. In various embodiments, the microcontroller 404a, SBC 404b, and 404c may comprise a single unit.

[0317] The microcontroller 404a may receive input from the knock sensor 428, and in response to the input signal received from the knock sensor 428, the microcontroller 404a may control or activate the servo motor 430 to open the access door to the magnet hole 402. The microcontroller 404a may also be configured to activate a single board computer (CPU) 404b in response to receiving an activation signal from the knock sensor 428. In at least some embodiments, the microcontroller 404a may be a low-power microcontroller, such as an Arduino board.

[0318] The single board computer (SBC) 404b, once activated by the microcontroller 404a, can control various electrical hardware components of the portable NMR device 104. For example, the SBC 404b can control the shimming control unit 422 to drive the shimming coil 424 to generate a compensation magnetic field that compensates for the spatial inhomogeneity in the magnetic field provided by the permanent magnet to provide a more uniform static magnetic field (B0). The SBC 404b can also include a communication module 432 (i.e., it can support IEEE 802.11 communication), and various input interfaces, including a universal serial bus (USB), a high-definition multimedia interface (HDMI), and an embedded multimedia controller (eMMC). In various cases, the USB interface can allow the NMR device to be used to send or receive data (i.e., as a data feed), or alternatively, as a power input for recharging the power module 412. In at least some embodiments, the SBC 404b can also include an onboard real-time clock (RTC). The RTC can be used to add a timestamp to the data measurement results generated by the portable NMR device 104.

[0319] The FPGA (digital signal processor) RAM 404 c may be configured to generate signals sent to the transmit unit 416 and receive signals generated by the receive unit 420 .

[0320] Still reference Figure 4B As described above, the transmitting unit 416 is configured to drive the transmitting coil 408a to generate a pulsed RF excitation field (B1). In various embodiments, the transmitting unit 416 is configured to drive the transmitting coil 408a to generate a circularly polarized or elliptically polarized magnetic field (B1) at one or more Larmor or resonant angular frequencies of the metabolites to be measured.

[0321] To this end, the transmitting unit 416 may include an in-phase transmission path 432 and a quadrature transmission path 434. The in-phase transmission path 432 may be configured to drive a first transmitting coil 436a in the transmitting coils 408a to generate an in-phase component of the RF excitation signal (B1). Similarly, the quadrature transmission path 434 may be configured to drive a second transmitting coil 436b in the transmitting coils 408a to generate a quadrature component of the RF excitation signal (B1). As explained in further detail herein, driving the transmitting coils 408a in-phase and in quadrature allows for the generation of circularly polarized or elliptically polarized excitation signals to induce resonance in the target nucleus.

[0322] Each of the in-phase and quadrature transmission paths 432, 434 includes a digital-to-analog converter (DAC) 432a, 434a coupled to a passive bandpass filter 432b, 434b through a broadband transformer or a balun 432c, 434c, respectively. The transformer 432c, 434c can provide galvanic isolation and impedance matching. The turns ratio of each transformer 432c, 434c can be selected to achieve optimal impedance matching. In various embodiments, the DAC 432a, 434a receives a digital signal generated by the processor unit 404 (or FPGA 404c) and converts the digital signal into an analog signal. The DAC 432a, 434a can also receive a sampling clock signal synchronized between the two DACs. Then, the analog signal generated by the DAC 432a, 434a is sent to the passive bandpass filter 432b, 434b across the transformer 432c, 434c. In particular, the bandpass filters 432b, 434b receive the signals generated by the DACs 432a, 434a and filter the signals for frequencies within a predetermined passband range of frequencies. In various cases, the passband range may include one or more Larmor (or resonant) frequencies for one or more target nuclear ensembles.

[0323] Similar to the transmit unit 416, the receive unit 420 includes an in-phase receive path 440 coupled to a first in-phase receive coil 438a, and a second quadrature receive path 442 coupled to a second quadrature receive coil 438b. As explained in further detail herein, the in-phase receive coil 438a is configured to receive (or detect) the in-phase component of the FID signal, while the second quadrature receive coil 438b is configured to receive the quadrature component of the FID signal. (See, for example, equations (6) and (7)). In at least some embodiments, the receive coil 438 can be the same as the transmit coil 436. In these cases, a fast PIN diode switch can be used to disconnect the receiver input in the transmit mode. In at least some cases, two PIN switches are used in series to improve isolation.

[0324] Each receive path 440, 442 includes a passive bandpass filter 440a, 442a coupled to each receive coil 438a, 438b at an input. Each bandpass filter 440a, 442a may have a similar topology as the bandpass filter 432b, 434b in the transmit unit 416. In accordance with the teachings herein, the bandpass filters 440a, 442a may be configured to filter a narrow frequency range including one or more Larmor frequencies of a target core.

[0325] The filtered signals produced by the bandpass filters 440a, 442a are then passed through wideband transformers and / or baluns 440b, 442b coupled to the outputs of the bandpass filters 440a, 442a, respectively. Similar to the transformers 432c, 434c of the transmit unit 416, the transformers 440b, 442b can provide galvanic isolation as well as impedance matching. In at least some embodiments, one end of the outgoing (or secondary) winding of each transformer / balun 440b, 442b can be center tapped for local ground reference. The turns ratio of each transformer 440b, 442b can be selected to achieve optimal impedance matching.

[0326] Each transformer 440b, 442b is coupled to a series connection at a node of the outgoing winding, which includes a low noise amplifier (LNA) 440c, 442c, a variable gain amplifier (VGA) 440d, 442d, a local oscillator (LO) 440e, 442e and an analog-to-digital (ADC) converter 440f, 442f. The LNA, VGA, LO and ADC are used to reduce signal noise and improve the FID or metabolite signal level to achieve higher measurement accuracy. In various embodiments, the combination of these components allows a low noise receiver design that achieves a noise level below 1.1nV / VHz (as reference input (RTI)).

[0327] In at least some embodiments, a clock generator, such as a phase-locked loop (PPL) clock generator 444, may be coupled to the LOs 440e, 440f and the ADCs 440f, 442f.

[0328] More specifically, LNA 440c, 442c is configured to function as a preamplifier stage to achieve an overall lower receiver noise figure. In at least some embodiments, LNA 440c, 442c is configured to produce an overall noise level of less than one decibel. LNA 440c, 442c may be formed, for example, from GaAs E-pHEMT technology.

[0329] VGAs 440d, 442d are coupled to the output nodes of LNAs 440c, 442c, respectively, and are configured to enhance the amplitude of the received FID signal. Specifically, VGAs 440d, 442d enhance the FID signal to mitigate parasitic signal products that would subsequently be introduced into the FID signal by the LO and ADC.

[0330] In at least some embodiments, the VGA 440d, 442d can be further incorporated into an automatic gain controller (AGC) circuit that is configured to provide gain control and limit the output of the VGA 440d, 442d. To this end, the AGC can use a feedback loop that resolves the difference between the output of the VGA 440d, 442d and an internal reference threshold. Incorporating the VGA into the AGC ensures that the LO mixer is not overloaded. For example, the LO mixer may be overloaded when insufficient water (or solvent) suppression is performed, so that the received FID signal includes not only the resonance signal of the glucose protons, but also the resonance signal of the hydrogen protons located in the water. In order to minimize the receiver noise figure over a large input level range, the AGC sets the maximum VGA gain without overloading the mixer RF input.

[0331] Coupled to the output of the VGA 440d, 442d is an LO mixer 440e, 442e, which is used to filter out other noise sources from the FID signal, including, for example, close-in LO phase noise, baseband 1 / f noise, and intermodulation distortion (IMD). In particular, the LO mixer 440e, 442e ensures that the noise in the received signal is spectrally uniform, has a Gaussian amplitude distribution, and is not corrupted by other noise signals. This is consistent with the assumptions used when calculating the statistical confidence level in equation (9). In at least some embodiments, the LO 440e, 442e is configured to shift the received FID signal from the Larmor frequency of the target proton by at least 100 kHz.

[0332] ADC 440f, 442f receives the analog output signal of LO mixer 440e, 442e, respectively, and converts it into a digital signal. In at least some embodiments, ADC 440f, 442f is a high-resolution 16-bit sampling ADC. For example, ADC 440f, 442f can be a 16-bit sigma-delta ADC converter, which is configured to oversample the analog output signal from LO mixer 4403, 442e at a rate greater than the Nyquist rate (i.e., for quantization). In various cases, in order to reduce quantization error and increase signal-to-noise ratio (SNR), "dithering" noise can be introduced into the received signal (e.g., white Gaussian noise). In at least some embodiments, "dithering" noise can be introduced to ensure that a non-harmonic spurious level of 110dB is achieved in the FID signal. In at least some cases, the ADC has an ultra-wideband analog input. In various embodiments, chemical shifts up to + / - 200 ppm in sub-2T systems can be supplied by low frequency or audio frequency sigma-delta converters with sampling speeds greater than 100 kSPS.

[0333] Reference now Figure 5A , a simplified circuit diagram of an example embodiment of a transmission circuit 500A is shown. The transmission circuit 500A may be Figure 4B 432, 434. In particular, the transmission circuit 500A shows the topology of the bandpass filters 432b, 434b in more detail. As shown, the transmission circuit 500A includes a digital-to-analog converter (DAC) 504, a transformer 506 (e.g., a transformer characterized by a turns ratio of N:1), a passive bandpass filter 508, and transmitting coils 510a, 510b.

[0334] As previously described, transformer 506 can be a broadband RF transformer for galvanic isolation and impedance matching. Transformer 506 can also be characterized by low insertion loss to minimize noise figure. In some cases, the primary winding of transformer 506 can include a center tap resistor 505, which is used for DC biasing of the DAC 504 output. In particular, the center tapped transformer can have an insertion loss of 0.5dB to 1.0dB. In at least some embodiments, the RF broadband transformer can be a transmission line type broadband balanced-unbalanced transformer with an insertion loss of less than 0.2dB.

[0335] The bandpass filter 508 may be similar to Figure 4B Any one of the bandpass filters 432b, 434b of the transmitting unit 416. As shown in the figure, the bandpass filter 508 can be a third-order or higher-order differential bandpass three-pass topology filter.

[0336] In particular, the bandpass filter 508 may include two parallel branches (ie, a first branch and a second branch), each of which includes an inductor (L S ) 508c, 508d are capacitors arranged in series (C S )508a, 508b. Each capacitor (C S ) 508a, 508b are coupled to the output node of the output (or secondary) winding of the transformer 506. The capacitor (C P )508e and inductor (L P ) 508f is arranged in parallel extending between the first branch and the second branch. Specifically, the capacitor (C P )508e and inductor (L P ) 508f has a common (or shared) node coupled to inductor 508c (L S ) of the output node, and the capacitor (C P )508e and inductor (L P ) 508f is coupled to the other common (or shared) node of the inductor (L S )The output node of 508d.

[0337] In the capacitor (C P )508e and inductor (L P ) 508f, is another set of parallel branches, each of which includes capacitors 508g, 508h (C S ). Each capacitor is 508g, 508h (C P ) has an output node that is serially coupled to an outbound coil winding 510a and a return coil winding 510b. In various embodiments, the outbound and return coil windings 510a, 510b correspond to the outbound and return coil windings of one of the first in-phase transmit coil 436a and the second orthogonal transmit coil 436b. Each outbound winding and return winding 510a, 510b can be located within the magnet bore and similarly polarized (i.e., in-phase or orthogonal). By implementing the transmit or receive coil as a half-length outbound winding and a half-length return winding, the coil length can be effectively reduced by a factor of two, resulting in an increase in the self-resonant frequency of the inductive coil, (i.e., the frequency limit at which the inductive coil loses its function and behaves as an open circuit). Increasing the self-resonant frequency of the inductor ensures that the self-resonant frequency is above the operating passband frequency.

[0338] A load resistor 512 is connected in series between the outbound winding 510a and the return winding 510b. Because the filter 508 provides a bandpass (voltage) amplitude frequency response across the load resistor 512, and the bandpass filter synthesized from the three-way topology low-pass filter has its final branch inductors 510a, 510b in series with the load resistor 512, the inductor current (i.e., the NMR coil current) also has the same amplitude frequency response characteristics. In addition, the outbound and return coil currents are directly related to the strength of the RF excitation field (B1) in each coil.

[0339] In various embodiments, the maximum voltage rating of the capacitor must be sufficient to cope with the operating voltage drop in resonance. In addition, the filter lumped inductance L P and L S They may be located outside the main aperture 106 field to avoid saturation of their ferrite cores by the static field (B0).

[0340] The bandpass filter topology described above avoids many of the problems associated with conventional transmit units. In particular, conventional transmit units in NMR applications employ narrowband, high-Q tuned LC circuits that use adjustable capacitors. The capacitors must be tuned to a given operating frequency (i.e., the Larmor or resonant frequency) to achieve impedance matching between the transmit coil (i.e., the load of the LC circuit) and the transmission line. Thus, the LC circuit is impedance matched for one operating frequency. Adjustable capacitors in conventional LC circuits are also unreliable, require time-consuming and expensive manual adjustments, and can introduce noise and gain errors.

[0341] In contrast, according to the teachings herein, a bandpass filter design does not include adjustable circuit elements and can be operated to filter a frequency passband that includes one or more Larmor frequencies rather than just one Larmor frequency. For example, this feature may support heteronuclear NMR experiments. For example, in one application, the emission passband range may include the hydrogen atoms in close proximity. 1 H and fluorine 19 F is the Larmor frequency. 19 The Larmor frequency of F is 1 The gain of the portable NMR device 104 is 0.94 times the Larmor frequency of H protons, a difference currently only supported by "double-tuned" resonant LC circuits. In particular, NMR studies of fluorine can be used to study the metabolism of fluorinated drugs and anesthetics. In addition, the bandpass tee topology described above provides a wider passband with a flat frequency response, allowing all spectral components to experience equal gain. This feature allows the portable NMR device 104 to operate over a wider Larmor frequency range without the need for retuning of the analog probe.

[0342] Allowing the transmission frequency to be within the passband (rather than at one operating frequency) also accommodates manufacturing tolerances in the permanent magnet design. The residual magnetization range of a permanent magnet typically varies by up to + / -3% around the nominal value. Since the Larmor frequency is a function of the static magnetic field (e.g., see equation (2)), variations in the residual magnetization from the nominal value result in variations in the Larmor frequency. Conventional LC circuits operating at a single frequency require a constant return frequency to accommodate manufacturing tolerances.

[0343] Reference now Figure 5B , a simplified circuit diagram of an example embodiment of a receiver circuit 500B is shown. The receiver circuit 500B may be Figure 4B 440, 442. As shown in the figure, the receiving circuit 500B includes a passive bandpass filter 508' and a transformer 506'. In particular, the bandpass filter 508' can be similar to Figure 4B The bandpass filters 440a, 442a of the receiving unit 420.

[0344] The bandpass filter 508' has Figure 5A The bandpass filter 508 of the embodiment of the present invention is of the same design as the bandpass filter 508 of the embodiment of the present invention. To this end, the bandpass filter 508' includes outbound and return coil windings 510a', 510b' corresponding to the outbound and return coil windings of one of the first in-phase receiving coil 438a and the second orthogonal receiving coil 438b. Each outbound winding 510a' and return winding 510b' can be located within the magnet bore and be similarly polarized (i.e., in-phase or orthogonal). A load resistor 512' is connected in series between the outbound winding 510a' and the return winding 510b'.

[0345] In at least some embodiments, transformer 506' may be center tapped for ground reference at the output (secondary winding). Transformer 506' may also be coupled to a low noise amplifier (LNA) 514, which may be similar to Figure 4B LNA 440c, 442c.

[0346] In various cases, the passband range of bandpass filter 508' may be different from the passband range of bandpass filter 508. This may allow for a passband range on the transmitter side that includes Larmor frequencies of one or more isotopes, while a passband range on the receiver side that detects only Larmor resonance signals of one specific isotope.

[0347] Reference now Figure 5C , showing the Figure 5A The bandpass filter 508 or Figure 5B500C shows an example frequency response graph of a bandpass filter 508'. Specifically, graph 500C shows a simulated frequency response for a third order Chebyshev response bandpass filter designed to support Larmor frequencies between 60 MHz and 80 MHz. Coil self-resonance is manifested as a stop band zero at approximately 200 MHz.

[0348] Reference now Figure 6 , shows a schematic diagram of an example coil assembly 600 according to some embodiments. One instance of the coil assembly 600 may be used to implement the RF transmit coil 408a, while a second instance of the coil assembly 600 may be used to implement the RF receive coil 408b.

[0349] The coil assembly 600 may include a first in-phase coil 600A and a second orthogonal coil 600B. In various embodiments, the coils 600A and 600B may be volume coils, such as solenoids or air inductors, which provide a more uniform RF excitation magnetic field (B1) across the magnet bore 106 at the transmit end and provide better receive sensitivity at the receive end (such as compared to surface coils currently used in high-field NMR applications). More specifically, in contrast to surface coils, volume coils are inherently capable of producing a more uniform excitation field that is independent of the use of adiabatic pulses. Adiabatic pulses that provide a uniform excitation field at certain thresholds are often used with surface coils to compensate for (or mitigate) the inherently non-uniform RF excitation fields that these surface coils typically produce. However, as previously described, because the adiabatic pulses must be truncated in the time domain, it may not be possible to use surface coils to produce high-frequency selective excitation pulses with high uniformity (e.g., as required in water suppression techniques). Therefore, the ability of volume coils to produce an inherently uniform excitation field that is independent of adiabatic pulses enables volume coils to be used in applications requiring high-frequency selective excitation pulses. Low-field designs of portable NMR devices (eg, generating fields below 2 T) are also adapted to the lower self-resonant frequency of the volume coil.

[0350] Still reference Figure 6 , the geometry of coils 600A and 600B may be a spiral coil geometry characterized by equation (20):

[0351] C(x, y, z)=C(R sinθcotα+pθ / 2π, R sinθ, R cosθ) (20)

[0352] Where, "R" is the radius of curvature of each coil, "Q" is the angle parameter, "α' is the tilt angle, and "p" is the pitch.

[0353] exist Figure 6, coils 600A, 600B are shown extending longitudinally in the x-direction and tilted relative to the YZ plane. When coil 600 is placed within the magnet bore, the z-axis of coil 600 defines the axis of the static magnetic field (B0).

[0354] In at least some embodiments, coils 600A, 600B may be spatially orthogonal tilted cosine coils that are tilted or canted in opposite directions and angled at ±45 degrees relative to the YZ plane to produce circularly polarized RF signals. In other cases, the spatial offset angle may be adjusted to generate, for example, elliptically polarized RF signals.

[0355] When coil 600A is used as a transmitting coil (e.g. Figure 4A 408A in FIG. 408A), a first in-phase coil is coupled to an in-phase transmission path 432 and is driven to generate an in-phase component of an excitation field (B1), while a second orthogonal coil 600B is coupled to an orthogonal transmission path 434 and is driven to generate an orthogonal component of an excitation field (B1). When the coils 600A, 600B are driven in in-phase and orthogonal directions, the coil 600 can generate a circularly polarized field (e.g., left-handed or right-handed, as required and defined by the gyromagnetic ratio of the target isotope), thereby inducing magnetic resonance in the target isotope. As previously described, the frequency of the transmitted excitation field (B1) can be controlled by a bandpass filter of the transmitting unit 416.

[0356] When the coil 600 is used as a receiving coil (eg Figure 4A 408B in the figure), the spatially orthogonal arrangement of the coils allows the coils to detect circularly polarized FID signals generated in the transverse XY plane. In particular, the first in-phase coil 600A can detect the in-phase component of the FID signal, while the second orthogonal coil 600B can detect the orthogonal component of the FID signal (e.g., see equations (4) and (5)). The sum signal and the difference signal of the two tilted cosine coils at the receiving end can be used to distinguish the magnetization signals projected onto the two transverse spatial axes (i.e., the X and Y axes), which may be necessary in some NMR spectroscopy applications.

[0357] Still reference Figure 6In order to support the coils 600A, 600B in a stable position, the coils 600A, 600B can cover a hollow cylindrical support structure 602. The support structure 602 can include a cavity 604 that can accommodate a subject's finger or a test sample when the structure is placed in the magnet hole 106. In at least some embodiments, the outer surface of the cylindrical support structure 602 can be lined with a spiral groove or notch that accommodates the spiral coil. The spiral groove or notch can help maintain the spiral form of the coil, as well as the position of the coils relative to each other. In particular, the mechanical support 602 and the notch or groove on its outer surface are important for offsetting the Lorentz force and maintaining the structural integrity of the coil geometry, which would otherwise deform the coil geometry over time. In at least some embodiments, the coil can include an insulating protective cover.

[0358] The cable connections between the coils and corresponding spectrometer components (e.g., transmitting unit 416 and receiving unit 420) can extend parallel to the axis of the static magnetic field (B0) (e.g., the z-axis) and be located within the magnet assembly frame of the portable NMR device 104 to avoid introducing any z-directed error components to the B0 field.

[0359] In various embodiments, the connection between the coils 600A, 600B and the spectrometer receive and transmit units 416, 420 is achieved via a shielded twisted pair cable. The shielded twisted pair cable can effectively prevent coupled noise and provide emission suppression, which is an important consideration considering the proximity of the operation of the portable NMR device 104 to the FM radio band. In at least some embodiments, the shielded twisted pair cable can be a Category 6A Ethernet cable, which contains four differential pairs of wide analog bandwidth (>500MHz) shielded twisted pairs and thus supports in-phase and quadrature signal delivery for transmit and receive operating modes.

[0360] Reference now Fig. 7A and 7B , shows an example embodiment of an axisymmetric permanent magnet assembly 700A according to some embodiments, which can be composed of Figure 1A The portable NMR device 104 is used. In particular, Fig. 7A Shown along Figure 1A A cross-sectional view of the portable NMR device 104 is shown along section line 7-7'. Figure 7B Shows Fig. 7A Exploded diagram of .

[0361] Fig. 7A and 7BThe axisymmetric permanent magnet assembly 700A shown in FIG. 7 is compact and is capable of producing a strong and nearly uniform static magnetic field (B0) with improved field confinement properties across the magnet bore 106. The compact size of the permanent magnet assembly 700A allows for a portable and consumer-friendly NMR device.

[0362] As shown therein, the axisymmetric design includes multiple magnet segments rotated around an axis of symmetry (e.g., z-axis). Each magnet is characterized in that the magnetization direction is either oriented along the axis of symmetry (z) or along the radial axis (r). As a result of the superposition of the magnetic fields generated by each permanent magnet segment, a static magnetic field (B0) is generated in the +z direction.

[0363] According to the teachings provided herein, a stronger, more uniform, and better confined B0 field is obtained because the Hilbert transform relationship between the orthogonal r and z magnetization components in each segment is realized over the entire range from 0 to 2π. In contrast, in conventional Halbach cylinder designs traditionally used for NMR applications, the Hilbert relationship between the r and φ components is only maintained for a limited length along the z direction.

[0364] In more detail, the axisymmetric permanent magnet includes a top disk magnet segment 702a and a top conical magnet segment 702b, which are stacked above the magnet hole 106 in the +z direction. Each of the top disk and top conical magnet segments 702a, 702b has a magnetization in the +z direction. In at least some embodiments, these segments may include a single magnet segment. The steel disk 716a can be inserted between the upper portion of the magnet hole 106 and the bottom surface of the top magnet disk 702b to enhance the static magnetic field (B0) and improve the uniformity of the static magnetic field. In particular, and by default, the magnetic field leaving the high permeability region (e.g., steel) and entering the low permeability region (e.g., air in the magnet hole 106) will leave perpendicular to the surface. Therefore, this feature allows for improved static field (B0) uniformity. In various embodiments, the surface of the steel disk 716a can be more accurately shaped to compensate for the inhomogeneity in the static magnetic field (B0) field.

[0365] Similarly, bottom disk magnet segment 704a and bottom cone magnet segment 704b are stacked below magnet aperture 106 in the -z direction. Each of bottom disk and bottom cone magnet segments 704a, 704b has a magnetization in the +z direction. Similar to segments 702a and 702b, in at least some embodiments, both segments 704a, 704b may also comprise a single magnet segment. A steel disk 716b may also be positioned between the bottom portion of magnet aperture 106 and the upper surface of bottom disk magnet segment 704b to enhance the static magnetic field (B0) and improve static magnetic field uniformity.

[0366] The central annular magnet segment 706 is positioned radially outward from the magnet aperture 106 in the +r direction and is characterized by being magnetized in the -z direction. The central annular magnet segment 706 at least partially surrounds the magnet aperture 106 while maintaining an access opening to the aperture 106 unobstructed (e.g., to receive a subject's finger or a test sample).

[0367] Stacked above the center annular segment 706 in the +z direction, and radially outward in the +r direction from the top disk and cone segments 702a, 702b is a top annular magnet segment 712 having a magnetization in the +r direction.

[0368] Similarly, stacked below the center ring segment 706 in the -z direction, and radially outward from the bottom disk and cone segments 704a, 704b, is a bottom ring magnet segment 710 having a magnetization in the -r direction.

[0369] In various embodiments, the top, center, and bottom permanent magnet rings may be implemented with multiple transversely magnetized cylindrical magnets with their magnetization oriented in either the +r or -r direction (eg, 12 magnets, 30 degrees apart).

[0370] As described above, the net magnetization intensity produced by each of these permanent magnet segments results in a static magnetic field (B0) in the vertical +z direction within the magnet bore 106. To this end, each magnet segment can be formed of a "hard" permanent magnet alloy such as a neodymium (NdFeB) permanent magnet, which allows the superposition principle to be applied. In at least some embodiments, N40 grade NdFeB permanent magnet material is used in the center ring 706 with a high reverse coercive field. In particular, the high reverse coercive field is generated by the strong static magnetic field (B0) in the magnet bore 104, which is tangential to and opposite to the magnetization of the center ring 706. Similarly, the top ring 710 and the bottom ring 712 can be formed of N40 grade NdFeB permanent magnets to provide optimal field confinement. N52 grade NdFeB permanent magnet material can be used for the top and bottom cone and disk segments 702 and 704. In various embodiments, modifying the rank of each magnet provides freedom when attempting to minimize the magnetic flux escaping from an ideal single sided flux structure, and therefore, allows for minimization of the size and weight of the protective cover surrounding the permanent magnet assembly.

[0371] Table 1 below summarizes the magnetization direction and material type of each magnet segment in the segmented axisymmetric magnet assembly 700A.

[0372] Table 1 - Axisymmetric magnet array design

[0373]

[0374] According to the teachings provided herein, the above-described segmented axisymmetric permanent magnet architecture ideally provides a nearly uniform and strong static magnetic field (B0) of 1.79 T across the magnet bore 106. In particular, this assumes that the permanent magnet structure has a height and diameter, each of which is approximately 10 cm (weight is approximately 10 kg). However, because the magnet bore 106 cuts into the center ring 706, the angular (or azimuthal) symmetry of the permanent magnet assembly is compromised and results in a reduction in magnetic field strength of approximately 4% to 6%, to, for example, approximately 1.68 T. However, a magnetic strength of 1.68 T represents an improvement over conventional segmented Halbach cylindrical permanent magnet architectures (commonly used in NMR) and normally achieves a static magnetic field strength of 1.54 T on the magnet isocenter line. As previously described, improving the static magnetic field strength in the magnet bore is an important consideration for improving the SNR of the received resonance signal in the portable NMR device 104 (e.g., the SNR is approximately equal to B0). 2 proportionally increased).

[0375] Fig. 7A and 7B The segmented magnet architecture also achieves a 5 Gauss field line limit. The 5 Gauss field line defines the outer perimeter of the magnetic field where the magnitude of the static magnetic field B0 is greater than 5 Gauss. The 5 Gauss field line limit is an important safety consideration for magnetic devices in regular commercial use.

[0376] Still refer to Fig. 7A and 7B , a hollow frame 718 can be disposed around the permanent magnet assembly to provide an outer 3D perimeter and an air gap around the axisymmetric permanent magnet. The hollow frame 718 can house various electrical hardware for operation of the NMR device 104. In various embodiments, the hollow frame 718 can be configured to receive Figure 4A and 4B 720 for analyzing the FID signal generated from the subject's finger or test sample located in the magnet bore 106). In various embodiments, with the coils 408a, 408b located in the magnet bore 106, the cable connection between the coils 408a, 408b and the spectrometer 720 can extend from the coils, through the magnet bore 106, down through the frame 718, and under the permanent magnet assembly to the spectrometer 720; or if the spectrometer 720 is provided externally, out the back of the permanent magnet assembly.

[0377] The hollow frame 718 can be formed of a non-magnetic material that protects the electrical hardware and circuits from the leakage magnetic field generated by the permanent magnet assembly. In at least some embodiments, the frame 718 can be formed of aluminum, which is both non-magnetic and lightweight.

[0378] To provide further insulation to prevent leakage of the electrostatic magnetic field, the NMR device 104 may also include an outer shielding layer 722 disposed around the hollow frame 718. The shielding layer 722 may be formed of any suitable material characterized by a high saturation field strength and a high magnetic permeability. In at least some embodiments, the material forming the shielding layer 722 may be characterized by a saturation field strength characteristic of greater than 1.5 T and a magnetic field strength characteristic of greater than 100 μm. o In at least some cases, the outer shielding layer 722 can be formed of a martensitic or ferritic stainless steel, which has high corrosion resistance and good magnetic properties (e.g., high magnetic permeability and high saturation strength). In various embodiments, the outer shielding layer 722 can be 3 mm thick. In some cases, an additional outer shielding layer can also be incorporated on the shielding layer 722 to provide additional leakage protection. The additional shielding layer 722 can be formed, for example, of a thin Mu metal.

[0379] Reference now Figure 7C and 7D , showing that Fig. 7A and 7B Simulation of the static magnetic field (B0) generated by a segmented axisymmetric permanent magnet assembly in . In particular, Figure 7C Shown by Fig. 7A and 7B A contoured field plot of the permanent magnet assembly produced. Fig.7D Graph 700D shows the strength of the static field generated within the magnet bore 106 as a function of radial distance from the magnet isocenter.

[0380] like Figure 7C As shown, the superposition of the magnetic fields generated by each of the magnet segments 702 to 712 is the magnetic field in the bore 106 along the axis of symmetry (+z). In addition, the axisymmetric arrangement of the permanent magnets combined with the outer shield 722 minimizes magnetic field leakage out of the device 104 and confines the static magnetic field therein.

[0381] Fig.7D A graph 700D of magnetic field strength is shown as a function of radial distance (i.e., along the transverse plane XY) from the magnet isocenter or the center of the magnet bore 106. As shown, the axisymmetric design achieves a static magnetic field strength of 1.79 Tesla at the isocenter, which tapers to only about 1.67 Tesla at 1.5 cm from the isocenter due to the loss of Azmitian axisymmetry caused by the magnet bore, as previously described.

[0382] Reference now Fig. 7E, a graph 700E showing a static magnetic simulation of the static magnetic field (B0) within the hole 106 is shown. In particular, the graph 700E shows the static field along the z-axis relative to the distance from the magnetic isocenter along the three Cartesian axes. As shown therein, the static magnetic field (B0) is almost uniform along the z-axis and the y-axis, but the uniformity along the x-axis is slightly impaired. Again, the non-uniformity along the x-axis is caused by the loss of magnet volume in the center ring 706 due to the hole 106.

[0383] Reference now Figure 7F , shows a permanent magnet assembly 700F of a portable NMR device 104 according to another example embodiment. In particular, the central ring segment 706' now includes an inner surface that is non-planar, which is defined by a hemispherical protrusion into the magnet bore 106. Shaping the inner surface of the central magnet segment 706' in this manner can mitigate the reduced magnetic field static magnetic field strength and spatial inhomogeneity of the static magnetic field due to volume loss due to the bore 106.

[0384] In various embodiments, the inner surface protrusions of the center ring segment 706' may be formed from a separate pole piece formed from high permeability mild steel or other magnetic alloys.

[0385] Reference now Figure 7G and 7H , showing the effect of deforming the inner surface of the center ring segment 706 ′, as shown in FIG. Figure 7F shown.

[0386] Figure 7G 700B shows a contour field graph of the permanent magnet assembly 700B. Figure 7C In comparison, the uniformity of the static magnetic field in the magnet hole 106 is improved.

[0387] Figure 7H A graph 700H is shown of magnetic field strength as a function of radial distance (i.e., along cross section XY) from the magnet isocenter or center of the magnet bore 106. In particular, curve 700H shows a significant improvement in field uniformity compared to curve 700D, especially within a distance of 1 cm from the isocenter.

[0388] like Fig.7D and 7H As shown in the graph in , while the axisymmetric permanent magnet design achieves a nearly uniform static magnetic field across the magnet bore, the static magnetic field is not completely uniform, especially at a distance far from the magnet isocenter. Therefore, in order to correct for the spatial inhomogeneity of the static magnetic field (B0), the portable NMR device 104 may include a shim assembly located in the magnet bore 106 (e.g., see Figure 4A and 4B424), which generates a compensating magnetic field to improve the uniformity of the static magnetic field. In particular, the shimming assembly allows the static field uniformity to be below 0.1 ppm, and ideally within 0.01 ppm.

[0389] Reference now Fig. 8A , shows an example shimming assembly 800A according to some embodiments. The shimming assembly 800A may be similar to Figure 4A and 4B The shim component 424 is provided.

[0390] As shown, the shim assembly 800A is formed of a plurality of current-carrying conductors 802a extending along the x-axis (as defined relative to the NMR device 104) and arranged in a circular configuration within or near the aperture 104 and spaced evenly circumferentially. Thus, the current-carrying conductors form a "cylindrical" shape in three dimensions. In various cases, the current-carrying conductors 802a can be positioned on a hollow cylindrical support frame 804a.

[0391] In particular, the shimming assembly 800A employs conductors 802a having a uniform linear density, wherein each current-carrying conductor 802a can receive an independently controlled direct current (DC), which can be provided by, for example, a dedicated digital-to-analog (DAC) converter (e.g., using a multi-channel DAC converter). In various embodiments, the DAC converter can be located at Figure 4A and 4B The shimming control unit 422 may be configured to provide a plurality of current generators (not shown) coupled to a plurality of digital-to-analog (DAC) converters. In particular, the use of uniform line density with varying currents is contrasted with conventional NMR shimming designs that employ complex wiring geometries and use correction coils with variable copper winding density.

[0392] As explained in further detail herein, the shim assembly 800A uses a sinusoidal distribution of DC current amplitude through a current carrying conductor that varies relative to the angular position of the current carrying conductor around a hypothetical "shell" of a "cylinder" (e.g., formed by the arrangement of the current carrying conductors). This feature, combined with the uniform circumferential spacing of the current carrying conductors, enables the shim assembly 800A to produce a high-order mode magnetic field pattern in the magnet bore 104, which in turn can accurately compensate for the inhomogeneities of the static magnetic field (B0) as described herein. In particular, conventional shim assemblies implement conductor coil geometries that are intended to produce a single magnetic field mode (when driven by a current of a magnitude proportional to that mode). Therefore, without a large number of complex coil wiring geometries for each mode, these shim assemblies are unable to provide accurate and precise compensation for the inhomogeneities of the static field.

[0393] Further, and as explained in detail herein, the ability to variably control the current through each linear conductor allows for a unique one-to-one mapping between the magnitude of the modal current and the magnitude of the corresponding term in the polynomial decomposition of the uncompensated static magnetic field (B0). That is, if the inhomogeneity of the static magnetic field (B0) is represented as a polynomial or decomposed into a polynomial (see, e.g., Fig. 7E , where the inhomogeneity curve graph can be approximated as an nth-order polynomial), the current in each conductor can be adjusted to compensate for the corresponding term of the polynomial. Significantly, this allows the shimming assembly to achieve accurate correction of a near-uniform static field. By extension, this allows the portable NMR device 104 to achieve high field uniformity on the order of 0.01 ppm (or at least less than 0.1 ppm), which is not achievable using conventional shimming assemblies. In particular, conventional shimming assemblies do not allow for similar direct one-to-one mapping and cannot promise accurate static magnetic field compensation to produce high static magnetic field uniformity useful in NMR applications.

[0394] It should also be understood that the present shim assembly 800A can be distinguished from conventional "bridge-cage coil" assemblies that may be used in conventional NMR applications to generate a uniform RF magnetic field (B1). In particular, a conventional birdcage RF coil has end cap rings that capacitively couple each "bar" of the bridge coil to the next bar. As a result, the birdcage is electrically a cascade of CL high-pass filters, with capacitors between the bars on the two end rings and the bars acting as distributed inductors. Thus, this bridge-cage topology allows a single feed at the correct frequency to establish resonance in the form of a sinusoidal distribution of current around the cage. Thus, the system provides a nearly homogeneous B1 field at the selected radio frequency and circular polarization. However, the bridge-cage topology cannot extend down to DC as is the case with the shim assembly 800A, and can only generate higher mode field patterns at harmonics of the tuned RF frequency. Therefore, by removing the end cap rings in the shim assembly 800A and driving each conductor line with independent DC currents generated by the multi-channel DAC, the shim assembly 800A can generate any magnetostatic mode field pattern in the magnet bore 104. As previously described, the ability to generate any mode field pattern allows for a high degree of correction for field inhomogeneities in the static magnetic field (B0).

[0395] Now in more detail, to generate a correction field using shim assembly 800A, the current density (I) in each current carrying conductor (also referred to herein as a "shim current pattern") varies sinusoidally. In at least some embodiments, this sinusoidal variation can be expressed according to Eq.

[0396] I=I o sinθ (21)

[0397] Among them, I ois the magnitude of the current flowing through the corresponding current carrying conductor, and θ is the angle of the current carrying conductor measured counterclockwise with respect to the z-axis.

[0398] Reference now Figure 8B , showing the Fig. 8A A cross-sectional view of the shim assembly 800A taken along the section line 8B-8B'. As shown by the field streamlines shown therein, where the current density varies according to equation (21), the superposition of the magnetic fields generated by each current carrying conductor generates a dipole field characterized by a linear compensation field along the z-axis. Figure 8B In the example shown in the figure, the conductor current density is 0.42MA / m 2 (million amperes per square meter) to 3.16MA / m 2 The linear compensation field along the z-axis can be used to compensate for the inhomogeneity of the static magnetic field (B0) along the z-axis accordingly.

[0399] In at least some other embodiments, the current density (I) in each current carrying conductor may also vary according to equation (22).

[0400] I=I o sin2θ (22)

[0401] The effect of varying the conductor current according to equation (22) is to produce a higher order quadruple mode field which is also characterized by a linear compensation field along the z-axis.

[0402] Reference now Figure 8C , showing that along Fig. 8A Another cross-sectional view of the shim assembly along the section line 8C-8C of FIG. As shown in the field streamline diagram, the current varies in the current-carrying conductor according to equation (22), a quadruple field is generated, and the compensation field along the z-axis is linear (although not shown, the field is also linear along the y-axis). Figure 8C In the example shown in the figure, the conductor current density is 0.82MA / m 2 (million amperes per square meter) to 3.08MA / m 2 varies within the range between.

[0403] Thus, shimming assembly 800A may be used to generate a 2n-pole field by varying the current density (I) in each current carrying conductor according to equation (23).

[0404] I=I o sin nθ (23)

[0405] As explained above (and as explained in further detail herein), allowing the generation of higher order mode fields allows for more accurate compensation of the inhomogeneous static magnetic field (B0). This is because in the higher order modes, the higher order terms can be related to the corresponding terms in the polynomial expansion of the static magnetic field (B0). Therefore, the shim coil assembly 800A allows a unique one-to-one mapping between the amplitude of each shim current mode and the amplitude of the corresponding term in the polynomial decomposition of the inhomogeneous static magnetic field (B0) (i.e., which is generated by the permanent magnet assembly).

[0406] The theory behind the unique one-to-one mapping of shim current patterns to corresponding terms in the polynomial decomposition of the inhomogeneous static magnetic field (B0) will now be described in more detail herein.

[0407] The magnetic field generated by the combination of the static magnetic field (B0) (ie, generated by the permanent magnet assembly) and the compensation field (ie, generated by the shim coils) satisfies the governing magnetic equation expressed in equation (24).

[0408]

[0409] in is the vector magnetic potential, It represents the current density through the current-carrying conductor in amperes per square meter, μ0 is the magnetic permeability of free space, represents the spatial magnetization of the permanent magnet structure. Given the high linearity of the hard permanent magnets used in the permanent magnet assembly of the NMR device 104, the superposition principle can be used to consider the effects of the shim current and the permanent magnets separately.

[0410] In particular, one can solve the vector magnetic potential by first Its warping is then solved to solve for the magnetic field (B) generated in the magnet bore 106 (ie, as a result of the static field and the compensation field). For the proposed current geometry, is purely longitudinal (x-axis) and satisfies the Laplace equation expressed by equation (25):

[0411]

[0412] Assume that the conductor extending along the x-axis has a radius r. i The cylindrical shape is assumed to be r = r on the "shell" of the cylinder. i The surface current is defined at a radius of , then the solution of the Laplace equation in equation (25) can be obtained by i The first region (region I) defined by r>r i is found in the defined second region (Region II), subject to the following continuity and normal derivative boundary conditions at r = ri, as provided by equations (26) and (27).

[0413] AJ (r i ,θ)=A II (r i ,θ) (26)

[0414]

[0415] Where θ is the angle relative to the z-axis, K no and K ne is the nth odd and even mode sheet current density in amperes per meter along the circumference of the cylinder, and N is the shim current mode.

[0416] It has been recognized that the surface current boundary conditions given by Fourier series based on circular harmonics are suitable for the longitudinal symmetry of this ideal 2D device geometry and can produce any desired compensating magnetic field (B) field distribution within the hole, such as when N (shim pattern) approaches infinity.

[0417] In particular, for an ideal distribution with a radius of r i Arbitrary current distribution of order N on an infinitely thin cylindrical shell, with r<r in the hole i The magnetic field (B) field can be expressed according to equation (28).

[0418]

[0419] Furthermore, the magnetic field (B) outside the cylindrical hole with r>ri can be expressed according to equation (29).

[0420]

[0421] At least from equation (28), it can be seen that the nth odd harmonic of the uniform current mode is related to the z-axis in the cylindrical hole. The magnitude of the tangential compensation magnetic field (B) is z (n-1) There is a direct correspondence between dependencies.

[0422] Therefore, this B0 field modal analysis shows a unique one-to-one mapping between the required amplitude of each shim current mode and the amplitude of the corresponding term in the polynomial decomposition of the tangential magnetic field (B0).

[0423] However, these shim currents are only approximations of the ideal surface currents which do not additionally take into account the spatial sampling at the shim angles and the finite current density in the cross section of each shim conductor. A shim structure with M conductors limits the polynomial correction order to N = M / 2-1. A good test of the theory is to relate the current in the lth conductor to the ideal surface currents and use the result to predict the static field (B0) strength. The circumference of the cylinder is divided into M regions spaced at uniform angles around each shim conductor. The current pattern n in the lth conductor is the line integral of the surface current in region l given by equation (30).

[0424]

[0425] In the above expression, only the odd-mode currents are retained (with the subscript o hanging down), because the static magnetic field (B0) along the x-axis comes from the odd-mode surface current distribution. For low-order modes in a shim structure with sufficient conductors, This simplifies equation (30) to equation (31).

[0426]

[0427] As mentioned before, the B0 dipole field in the hole is associated with mode n=1, so equation (29) resolves to equation (32).

[0428]

[0429] The model predicts that mode 1 (θ k =π / 2) produces a 0.16mT B0 field at the isocenter of the magnet with a 1A peak current, a 1.5cm aperture radius and 24 shim conductors. Although it is common practice to take the B0 field at the isocenter of the magnet as the "true" value and shim only n>2 higher order modes, the n=1 modal analysis demonstrates the relative strength of the shim system and provides a good check on the theory as it agrees well with the empirical results.

[0430] In practice, each shim current pattern generates a compensation field that is best described by an Nth-order polynomial with a dominant term, although it has several non-negligible additional terms. Let S ij represents the polynomial fit to the field produced by the jth current mode with all other shim currents turned off. i The coefficients of the terms. These field response polynomials can be determined experimentally for each shimming current mode. If the uncompensated static field (B0) is written as an Nth order polynomial represented by equation (33), the current required to shim the system can be found by solving equation (34).

[0431] B0(z)=B0+ΔB 01 z+ΔB 02 z 2 ...ΔB 0N z N (33)

[0432]

[0433] The elements of the current column vector are I j＝k-1 Modal current.

[0434] It can be observed that the matrix describing this linear system is strongly diagonal and therefore well-conditioned and invertible. Conventional shim coil geometries (which do not directly correspond to the system modal properties) may result in a poorly conditioned matrix formulation requiring the computation of a pseudo-inverse. In this case, the field homogenization performance may suffer.

[0435] In various cases, the method of resolving the shim currents using equations (33) and (34) can be applied when a shim assembly is used with a conventional Halbach magnet array design defined by longitudinal symmetry.

[0436] Reference now Fig.8D , shows a shim assembly 800D according to another example embodiment. In particular, the shim assembly 800D may represent Fig. 8A The embodiment of the shim assembly 800A is more suitable for Fig. 7A Or the axisymmetric permanent magnet design shown in FIG7F. The theoretical principle underlying the shim assembly 800D is similar to the theoretical principle explained above with respect to the shim assembly 800A.

[0437] As shown, the shim assembly 800D includes two sets of parallel, laterally spaced concentric rings 801A, 801B, which may be located on the top and bottom surfaces of the cavity 106 such that the centers of the concentric rings are aligned with the axis 106 (e.g., Fig. 7A and 7F The z-axis in the static magnetic field is colinear. Each group includes a plurality of concentric rings that rotate around the axis of the static magnetic field. In particular, the concentric rings 801A, 801B on the top surface and the bottom surface respectively form a plurality of spaced and relative concentric ring pairs 802a, 804a, 806a, 808a and 810a and 802b, 804b, 806b, 808b and 810b. In various cases, at least one current-carrying conductor pairing can be configured so that the coil radius of each coil in the conductor pairing is substantially equal to the spacing between the corresponding coils of the conductor pairing. Therefore, if the current drive directions of the two coils are opposite (e.g., clockwise and counterclockwise), this can satisfy the inverse Helmholtz condition, which allows the coil pairing to generate a linear field gradient between the two coils.

[0438] In various embodiments, each central ring carries an independent current and produces a field distribution on an axis of symmetry (e.g., the z-axis). Although the magnetic field distribution is known analytically in closed form in a manner similar to that discussed above with respect to shimming assembly 800A, a polynomial expansion of the magnetic field produced by each coil using spherical harmonics may be better for evaluating the suitability of concentric coils 800D for uniform compensation of the static field (B0).

[0439] In particular, the magnetic field components generated by the circular shim coils in assembly 800D, whose radius (a) is positioned at a polar angle (α) relative to the magnet isocenter, can be expressed in the form of spherical harmonics using equations (35) and (36). Equation (35) represents the radial (r) direction component of the field, while equation (36) represents the polar angle (θ) direction component. Equations (35) and (36) apply to the region within the radius of the shim coils (e.g., r>a).

[0440]

[0441]

[0442] Where I is the current applied to the shim coil (A), P n is the Legendre polynomial, is the associated Legendre polynomial of the first kind (WRSymthe[3]).

[0443] exist Fig.8D In an exemplary embodiment of the present invention, each coil in groups 801A and 801B is oriented relative to the isocenter of the magnet by α=tan -1a / h defined by the polar angle (α) positioning. Where ±h is the coil height above or below the plane z = 0. For permanent magnet structures (e.g., Figure 7F The magnet 700F) is ideally axisymmetric in design, with no azimuthal magnetic field component or no dependence on Furthermore, on the axis of symmetry, where cosθ = 1, the field is purely radially directed (r = z), which is true for all values ​​of n. and P n (1)=1.

[0444] Each closed circular coil in coil groups 801A and 801B generates even and odd Legendre polynomial terms in equations (35) and (36), rather than a single dominant mode. In addition, coils of different radii generate linearly independent Legendre polynomial coefficient vectors. In particular, this is due to the fact that the Legendre polynomial coefficients depend on the power of the coil radius (a). In particular, by using an infinite series expansion on spherical harmonics to uniquely represent the magnetic field generated by the shim coils, and demonstrating that each coil generates even and odd polynomial terms (rather than just a single dominant mode), it can be understood that the shim assembly can be used to form a complete basis to generate any desired compensation field (e.g., any desired field pattern) in the hole 106 of the axisymmetric permanent magnet assembly defined by azimuthal symmetry.

[0445] It will also be appreciated that the separability of the magnetic field equations in the radial (r) and polar (θ) directions means that correction of the field to make it uniform along an axis that is collinear with the magnetic field (e.g., the z-axis in a 7F magnet) will ensure that the magnetic field will be uniform everywhere within the magnet bore. In particular, this fact results from the axisymmetric design of the permanent magnets and greatly simplifies the shimming process.

[0446] Reference now Fig.11 , shows an example process flow for a method 1100 for calculating the current of each shim coil in a shim assembly 800D to correct an axisymmetric magnetic design (eg, Figure 7F The inhomogeneity of the static magnetic field (B0) in the magnet design 700F).

[0447] In act 1102, the static magnetic field (B0) is mapped when all shim currents are set to zero. A field map of the uncompensated static magnetic field (B0) along the z-axis is then generated by a gradient imaging technique (e.g., Figure 7F ). In various cases, gradient imaging techniques use a linear one-dimensional gradient applied along a static field axis (e.g., the z-axis). This can be accomplished, for example, using a magnet located inside the bore of the magnet. Fig.8D The same shim device 800D of FIG. 800A may be used, or alternatively, a shim device 800D may be used that is oriented orthogonally to the static magnetic field axis (e.g., along the x-axis) in the magnet bore. Fig. 8A The coil assembly 800A is used to generate a linear gradient. The gradient imaging technique is generated using a test sample (e.g., a water sample) of uniform proton density located in the hole. In various cases, the k-space dimension can also be converted to a field sampled along the gradient axis using a discrete Fourier transform (DFT) pairing. In some embodiments, the field profile can be expressed as a P-order polynomial, as shown in equation (37).

[0448] B0(z)=B0+ΔB 01 z+ΔB 02 z 2 ...ΔB0N z P (37)

[0449] In other embodiments, the static field profile (B0) along the z-axis can be represented according to a column vector. The column vector can be an Mx1 vector representing the uncompensated static field (B0(z)) along the z-axis, where M is the number of equally spaced points sampled along the z-axis. The static field inhomogeneity error (ΔB0(z)) can then be determined according to equation (38).

[0450] ΔB0(z)=B0(z)-B0(0) (38)

[0451] where B0(0) is the vector of the uniform compensated static field along the z-axis.

[0452] At act 1104, a unit current is applied to each coil in coil assembly 800D, and the resulting magnetic field of each coil is mapped. In subsequent iterations of method 1100, as described in further detail herein, an incremental current may be applied to each shim coil in addition to the shim current value determined at act 1108 in a previous iteration of method 1100. Similar to shim assembly 800A, the current applied to the concentric rings may be determined by Figure 4A and 4B The current shimming control unit 422 applies. To this end, the shimming control unit 422 may include a plurality of current generators (not shown) and a plurality of digital-to-analog (DAC) converters, wherein each DAC is coupled to one of the current generators and one of the concentric rings to provide a unique current thereto.

[0453] In action 1106, the static magnetic field (B0) profile determined in action 1102 is subtracted from each magnetic field determined in action 1104 for each coil. In the first iteration of method 1100, this allows determination of the static magnetic field generated by applying a unit current to each coil. In particular, for the jth coil, the static magnetic field (B0) profile can be determined based on the Mx1-dimensional column vector (S j ) represents the magnetic field profile along the z-axis (eg, the static field axis). In subsequent iterations of method 1100, action 1106 allows for determining the static field generated by applying the incremental current, which is in addition to the shim current values ​​determined in the previous iteration of method 1100.

[0454] At act 1108, the shim current required for each coil to produce a uniform static magnetic field in the magnet bore is determined.

[0455] In particular, the static bore magnetic field generated by the combination of coil fields determined in act 1106 can be assumed to be a linear superposition of the magnetic fields generated by each shim coil. Thus, the current required to shim the spatial inhomogeneities in the static magnetic field (B0) can be determined from the subsequent linear system represented by equation (39).

[0456]

[0457] The shim matrix S is rectangular because the number of points sampled along the z-axis (M) is greater than the number of shim coils (N).

[0458] The shim currents required to produce a uniform static magnetic field in the magnet bore can be determined by solving the linear system represented in equation (39). This can be done by inverting the product of the shim matrix and its transpose, as shown in equation (40).

[0459] I=(S T S) -1 S T (-ΔB0) (40)

[0460] Then, at 1110, the current determined for each shim coil in equation (40) can be applied to the shim assembly 800D to correct for inhomogeneities in the static magnetic field (B0) inside the magnetic bore. For example, as previously explained, this can be implemented using a shim control unit 422 that includes a plurality of current generators (not shown) and a plurality of digital-to-analog (DAC) converters, wherein each DAC is coupled to one of the current generators and one of the concentric rings to provide a unique current thereto.

[0461] In various cases, method 1110 may require multiple iterations to resolve the shim currents that result in a uniform static field (B0). These iterations are caused by nonlinearities in the permanent magnet B(H) characteristics, which violate the assumption of linear superposition in equation (39). In particular, the intermediate ring magnet (e.g., Fig. 7A 706 or Figure 7F 706' in action 1102) typically operates with a high reverse coercive field in its nonlinear B(H) region. Therefore, at 1112, it can be determined whether the shimming current applied at 1110 is sufficient to provide a threshold uniformity (e.g., less than 0.1 ppm) for the static field. For example, this determination can be made by mapping the static field generated by the shimming component in a manner similar to that described in action 1102. However, in contrast to action 1102, rather than configuring the static field by setting the shimming current to zero, the static field is configured by applying the shimming current determined at action 1108.

[0462] If the shim current is not determined to be sufficient at act 1112, method 1100 may return to act 1104 and repeat. As previously described, in subsequent iterations of method 1100, an incremental current (eg, which is incremental relative to the current determined at act 1108 in the previous iteration) is applied to each coil.

[0463] In other cases, if it is determined in act 1112 that the shim current is sufficient, then method 1100 may be further determined to be completed in act 1114 .

[0464] Because the permanent magnet assembly 700A and / or 700F generates a more spatially uniform static magnetic field (B0) than a conventional Halbach magnet array, the shimming assembly 800D generally requires less current (I) in order to provide sufficient magnetic field correction to achieve field uniformity less than 0.1 ppm. At least in some cases, the required current can be provided by a small portable power source (e.g., a lithium-ion battery) that can be easily placed within the frame 718 of the portable NMR device 104. This is in contrast to current compact permanent magnet designs, which require large magnetomotive forces measured in hundreds or thousands of ampere-turns to correct for inhomogeneities in the static magnetic field. The power supplies required to provide these current levels are incompatible with small portable NMR devices.

[0465] Reference now Fig. 8E , shows an example graph 800E of simulation results that plots the field along the field axis in the bore of an axisymmetric magnet (e.g., Figure 7F The profile of the uncompensated static magnetic field (B0) in the z-axis of the bore of magnet 700F is compared before applying shim current to shim assembly 800D and after the first iteration of method 1100. Fig. 8E The simulation results assume that each of the first and second coil groups 801A and 801B of the coil assembly 800D includes eight coils. As shown, a single iteration of the method 1100 can generate substantially greater static magnetic field (B0) uniformity in the magnet bore along the field axis.

[0466] Reference now Figure 8F , shows an example graph 800F, which shows Fig. 8E 1100 and is overlaid with a high-order (32nd order) polynomial fit. The RMS error of the first shimming result is 15 ppm. It should be understood that through further iterations of method 1100, the RMS error can be reduced.

[0467] Reference now Figure 8G , showing the use of Figure 8F800G is a graph of the residual error of the high-order polynomial fit. As shown, the high-order polynomial fit reduces the RMS error of the field to 0.56 ppm. As previously described, the RMS error can be reduced by further iterations of method 1100.

[0468] Reference now Figures 8H-8J , showing different variations of shim coil configurations. In particular, Figure 8H and 8J The coil configuration shown in can be used to correct for azimuthal variations due to spatial variations in the magnetization intensity and / or magnetization direction of the permanent magnet. Correcting for azimuthal variations can allow method 1100 to be more effective in homogenizing the field in the "r" (radial) and "theta" directions.

[0469] Figure 8I A coil configuration 800I is shown, which is Fig.8D The configuration of coil assembly 800D is similar to that of FIG. In particular, the configuration includes nested concentric coils designed to correct the fields in the "theta" and "r" (radial) dimensions.

[0470] Figure 8H An alternative example configuration 800H is shown in which twenty-four shim coils are provided at an offset of 15 degrees, and thereby, the coil diameter 802H is equal to one-third of the diameter 804H of the disk configuration formed by the combined coil group. The current feed and return are at right angles to the coil plane. In various cases, the configuration 800H can be used to correct the azimuthal variation of the magnetic field. In other cases, the shim coils can be offset at an angle less than 15 degrees to provide higher resolution for field correction. In other cases, higher angular offsets can also be used to simplify manufacturing. In some cases, the offset angle can be in the range of 10 degrees to 45 degrees. In addition, more or less than twenty-four coils can be included in each coil group 801A and 801B, and the coils can have a coil diameter 802H less than or equal to the disk diameter 804H.

[0471] Figure 8J Another example configuration 800J is shown where twenty-four shim coils are positioned with a 15 degree offset and the coil diameter 802J is equal to the disk radius 804J. In various cases, the configuration 800J can also be used to correct for azimuthal variations of the field.

[0472] In particular, in each of configurations 800H and 800J, the correction for the field in the azimuth direction is that the center of each coil is positioned at a unique angular coordinate Therefore, when the current in each coil is different, the static magnetic field in the z direction (B0) will depend on the azimuthal coordinate It should also be understood that although coil configurations 800H and 800J are designed to produce field corrections that vary with azimuth, these configurations also produce field dependencies in radial and polar coordinates. In contrast, coil configuration 800I produces a field that has no azimuth dependency (given its symmetry). Therefore, in the case of using coil configurations 800H and 800J for shimming, first using the coils to compensate for the inhomogeneities in the azimuth direction, and then using coil configuration 800I along the static magnetic field axis (e.g., the z-axis) One-dimensional gradient shimming can complete the process.

[0473] Reference now Fig. 9 , shows an example block diagram of a power and control system 900 for a portable NMR device.

[0474] As shown, the power and control system 900 includes a gradient shimming module (GSM) 902 and a spectrometer board 904. The gradient shimming module 902 can be used to provide a gradient shimming coil (e.g., Figure 4B The shim coil 424 or Fig. 8A The shim coil 800a or Fig.8D The spectrometer board 904 can be used to power a spectrometer (e.g., Figure 4B In various cases, each of the GSM 902 and the spectrometer board 904 may include a circuit board, such as a printed circuit board (PCB).

[0475] GSM 902 may receive power from power module 906. Power module 906 may be similar to Figure 4AThe power module 906 may be a compact battery configured to be housed within the portable NMR device 104 (e.g., a lithium ion (Li-Ion) battery), or may be an AC-DC power adapter to be powered from an external AC power source, as previously described. In various cases, power may be delivered from the power module 906 to the GSM 902 via a USB-C interface. For example, the power module 906 may include a USB port 903a that may communicate with a USB port 903b of the GSM 902 via a USB cable 903. In at least some cases, the USB cable 903 may be a USB-C cable such that power is delivered from the power source 906 to the GSM 902 via the USB-C interface. Thus, the GSM 902 may include a port 910 that complies with the USB-C Power Delivery protocol. In particular, the USB-C port may allow the system to negotiate a power contract with a USB-C compliant AC power adapter or battery pack to obtain only the required power. Generally, a power contract occurs when both ends of a USB Type-C cable agree on a power level to be delivered from a power source 906 to a powered device (e.g., GSM 902). In various cases, this can save power by avoiding power dissipation that might otherwise be necessary to accommodate worst-case magnet inhomogeneities and associated shimming currents.

[0476] In at least some cases, the GSM 902 may also include a DC-DC converter 912 for powering the shimming coils. Since the shimming strength is proportional to the current, rather than the power, the use of the DC-DC converter 912 may allow the USB-C maximum power delivery of 100W (20V, 5A) to be converted to a lower voltage (e.g., 5V, 20A) with a higher current capacity. In various cases, since the DC-DC converter 912 may generate switching noise that may degrade the shimming, the GSM 902 may also include a notch filter 914 to reduce the noise. The output of the DC-DC converter 912 is coupled to the input of the notch filter 914. The output of the notch filter 914 may then be used to power a coil driver 916, which may include a non-volatile memory random access (NVRAM) memory and a DAC coupled to the shimming coils of the shimming assembly. The NVRAM may store shimming current values ​​(e.g., determined using the method 1100) for each shimming coil after the NMR system is shut down. This therefore avoids the need to re-determine the appropriate shim values ​​each time the system is powered on. In various cases, the number of DACs may be equal to the number of shim coils. Alternatively, a multi-channel DAC having D channels may be used, where D is the number of shim coils.

[0477] In at least some embodiments, the connection between the GSM 902 and the spectrometer board 904 can occur through a USB interface 920 (e.g., a USB-C interface), which provides power from the GSM 902 to the spectrometer 904 and also sends control data from the spectrometer board 904 to the GSM 902. In some cases, a USB 2.0 serial bus embedded in the USB-C interface can be used to send control and data information. The GSM 902 can also include a USB2 to Serial Peripheral Interface (SPI) bridge 918 for relaying control signals received from the spectrometer board 904 to the coil driver 916 via the USB interface 920. In various cases, the separation between the GSM 902 and the spectrometer board 904 facilitates system integration and packaging, and meets different design rules and regulatory requirements for high-power DC and sensitive RF functions.

[0478] In at least some cases, the spectrometer board 904 may also include a USB port 922 (eg, a USB-C port) that may be used to connect the spectrometer board 904 to an external computer to receive control data or to send collected data.

[0479] Although the applicant's teachings described herein are combined with various embodiments for illustrative purposes, it is not intended that the applicant's teachings be limited to such embodiments, as the embodiments described herein are intended as examples. On the contrary, the applicant's teachings described and illustrated herein encompass various alternatives, modifications, and equivalents without departing from the embodiments described herein, the overall scope of which is defined by the appended claims.

[0480] References

[0481] [1]JCBoyd and DEBruns, "Quality Specifications for Glucose Meters:Assessment by Simulation Modeling of Errors in," Clinical Chemistry, vol.47, no.2, pp.209-214, 2001.

[0482] [2]RAdGraaf, in vivo NMR Spectroscopy, John Wiley & Sons Ltd., 2007.

[0483] [3]WRSymthe, "Static and Dynamic Electricity", 2nd edition, p.275, McGraw Hill.

Claims

1. A method for in vivo and non-invasive quantification, or in vitro and non-invasive quantification of the concentration of small molecule metabolites in a sample using a portable nuclear magnetic resonance (NMR) device, the method comprising: Inducing magnetization of the sample by applying a near-space uniform static magnetic field (B0) using a magnet accessory; suppressing a water signal generated by magnetization of water in the sample, the sample comprising a glucose hydrogen proton ensemble, wherein suppressing the water signal comprises using a radio frequency (RF) transmission unit for: applying a first RF excitation pulse (B1) to rotate the water magnetization into the transverse plane, wherein the first RF excitation pulse (B1) has an amplitude frequency response that reduces the response to one or more chemical shift resonance frequencies associated with one or more of the glucose hydrogen proton ensembles located in the sample, wherein one or more of the glucose hydrogen proton ensembles has a chemical shift that coincides with a frequency zero of the amplitude frequency response of the applied first RF excitation pulse (B1), wherein the amplitude frequency response is given by the expression |sinc n (πΔfτ)| definition, where n is an integer and n=1 or n=2, and nτ is the duration of the pulse, arranged to locate the zero point of the sinc function on at least one of the chemical shifts, and Δf is the offset frequency of the chemical shift resonance frequency from the water signal resonance center frequency, and; applying a dephasing pulse gradient to the sample to dephase the coherence of the spin precession of water in the sample after applying the first RF excitation pulse (B1), and wherein the first RF excitation pulse (B1) and the dephasing pulse gradient suppress a water signal generated by the magnetization of the water in the sample; After applying the dephasing pulse gradient, applying a second RF excitation pulse (B1) to the sample using the RF transmitting unit to excite one or more glucose hydrogen proton ensembles; detecting a free induction decay (FID) signal generated by the sample by using an RF receiving unit; converting the FID signal into a magnetic resonance spectrum using a processor unit; and The concentration of a metabolite in the sample is determined using the processor unit based on one or more resonance peaks in the magnetic resonance spectrum, the one or more resonance peaks having a chemical shift coinciding with a frequency zero point of an amplitude frequency response of an applied first RF excitation pulse (B1).

2. The method according to claim 1, wherein τ is defined as τ = (f H2O -f β-glc ) -1 , where f H2O is the resonant frequency of water protons, f β-glc is the resonance frequency of the β isomer, so that the frequency zero of the amplitude frequency response of the applied first RF excitation pulse (B1) coincides with the chemical shift in the magnetic resonance spectrum, and the method comprises: Prior to converting the FID signal into the magnetic resonance spectrum, homonuclear decoupling is applied to the sample by applying a low continuous wave radiation pulse, wherein the homonuclear decoupling is applied at least partially simultaneously with detecting the FID signal.

3. The method according to claim 1, wherein: The quantification of the concentration of the small molecule metabolite includes the quantification of the glucose concentration, wherein: Applying the first RF excitation pulse (B1) to rotate the water magnetization into the transverse plane is performed such that the first RF excitation pulse has an amplitude frequency response that reduces the excitation of the chemical shift resonance frequency associated with the glucose hydrogen proton ensemble including alpha and beta group hydrogen protons, the amplitude frequency response being given by the expression |sinc n (πΔfτ)| definition, where n is an integer and n=1 or n=2, nτ is the duration of the pulse, arranged to locate the zero of the sinc function at least on the chemical shift of the β-group hydrogen protons, and Δf is the offset frequency from the resonant center frequency of the water signal; After applying the dephasing pulse gradient, applying the second RF excitation pulse (B1) is performed to excite the glucose hydrogen proton ensemble including the α and β group hydrogen protons; The free induction decay (FID) signal is generated by relaxation of the glucose hydrogen proton ensemble; and The method includes determining, using the processor unit, a glucose concentration in the sample based on resonance peaks in a magnetic resonance spectrum generated by the glucose hydrogen proton ensemble.

4. The method according to any one of claims 1 to 3, further comprising: A discrete Fourier transform (DFT) is applied to the FID signal using the processor unit to convert the FID signal into a magnetic resonance spectrum, wherein a resonance occupied bandwidth is minimized by performing homonuclear decoupling.

5. The method according to any one of claims 1 to 3, wherein: The static magnetic field (B0) has a magnitude between 1.5 Tesla and 2 Tesla.

6. The method according to claim 3, wherein: The static magnetic field (B0) has a field uniformity between 0.01 ppm and 0.1 ppm.

7. The method according to any one of claims 1 to 3, wherein: The NMR device is suitable for use in a home environment.

8. The method according to claim 2, wherein: The method includes generating a static magnetic field (B0) of 1.68 Tesla, and the resonant frequency (f H2O ) is 71.5 MHz, and τ is 163 milliseconds.

9. The method according to claim 3, wherein: The method includes providing the first RF excitation pulses (B1) by generating an envelope modulated pulse train having triangular excitation pulses.

10. The method according to claim 3, wherein: The dephased pulse gradients are generated by a set of DC coupled shim coils.

11. The method according to claim 3, wherein: The method includes generating a first RF excitation pulse (B1) to be left-hand circularly polarized.

12. The method according to claim 11, wherein: The method includes applying the first RF excitation pulse (B1) for less than 1.5 ms to mitigate transverse relaxation decay of the glucose hydrogen proton ensemble at an exponential rate T2*.

13. The method according to claim 3, wherein: Prior to detecting the FID signal, the method includes applying spectral editing to distinguish the glucose resonances from those generated by other macromolecules.

14. The method according to claim 3, wherein: τ is defined as τ = (f H2O -f β-glc ) -1 , where f H2O is the resonance frequency of water protons, f β-glc β-glucose anomer 1 The method further comprises: before converting the FID signal into the magnetic resonance spectrum, by 2 Homonuclear decoupling is applied to the glucose hydrogen proton ensemble by applying a continuous wave radiation pulse at a resonance frequency of CH hydrogen radical protons, wherein the homonuclear decoupling is applied at least partially simultaneously with detecting the FID signal.

15. The method according to claim 14, further comprising: in, The glucose hydrogen proton ensemble includes α-glucose anomer 1 CH hydrogen group proton and β-glucose anomer 1 CH hydrogen group proton.

16. The method according to claim 15, wherein: The method comprises generating the second RF excitation pulse (B1) within a frequency range including the α-glucose anomer 1 CH hydrogen group proton and β-glucose anomer 1 Chemical shift of the protons of the CH hydrogen group.

17. The method according to claim 15, wherein: The method comprises the steps of: 1 CH hydrogen group proton and β-glucose anomer 1 The continuous wave radiation pulses are applied at the resonant frequency of the CH hydrogen radical protons.

18. The method according to claim 15, wherein: The method comprises: using the processor unit, based on the α-glucose anomer in the resonance spectrum 1 CH hydrogen group proton and β-glucose anomer 1 A one-to-one mapping of the amplitudes of the resonance peaks of the CH hydrogen group protons is used to determine the glucose concentration in the sample.

19. The method of claim 18, wherein determining the glucose concentration in the sample comprises: The α-glucose anomer 1 CH hydrogen group proton and the β-glucose anomer 1 The amplitude of the resonance peak of the CH hydrogen group protons is correlated with a known glucose concentration reference standard.

20. The method of claim 18, wherein determining the glucose concentration in the sample comprises: Determine the α-glucose anomer 1 CH hydrogen group proton and the β-glucose anomer 1 Anomer ratio of the resonance peak of the CH hydrogen group proton.

21. The method according to claim 18, wherein: Determining the glucose concentration in the sample includes determining the anomeric ratio and the temperature and pH level of the sample.

22. A compact portable NMR device comprising: a housing that provides magnetic shielding; a permanent magnet assembly disposed within the housing and configured to apply a near-uniform magnetic field (B0) to induce magnetization of the sample; A radio frequency (RF) transmitting unit, wherein the radio frequency (RF) transmitting unit is configured as follows: applying a first RF excitation pulse (B1) to rotate the magnetization of water located in the sample into the transverse plane, wherein the first RF excitation pulse (B1) has an amplitude frequency response that reduces the response to one or more chemical shift resonance frequencies associated with one or more glucose hydrogen proton ensembles located in the sample, the one or more glucose hydrogen proton ensembles having a chemical shift that coincides with a frequency zero of the amplitude frequency response of the applied first RF excitation pulse (B1), the amplitude frequency response being given by the expression |sinc n (πΔfτ)| is defined, where n is an integer and n=1 or n=2, nτ is the duration of the pulse, arranged to locate the zero of the sinc function on the at least one chemical shift, and Δf is the offset frequency from the resonant center frequency of the water signal, and; After applying the first RF excitation pulse (B1), applying a dephasing pulse gradient to the water sample to dephase the coherence of the spin precession of water in the sample, and wherein the first RF excitation pulse (B1) and the dephasing pulse gradient suppress a water signal generated by magnetization of water in the sample, the sample also comprising a glucose hydrogen proton ensemble, and After applying the dephasing pulse gradient, applying a second RF excitation pulse (B1) to the sample to excite at least a subset of the glucose hydrogen proton ensemble, an RF receiving unit configured to detect a free induction decay (FID) signal generated by relaxation of one or more glucose hydrogen proton ensembles; and a processor unit, the processor unit being coupled to the RF receiving unit, the processor unit being configured to: converting the FID signal into a magnetic resonance spectrum; and The concentration of the metabolite in the sample is determined based on one or more resonance peaks in the magnetic resonance spectrum, the one or more resonance peaks having a chemical shift coinciding with a frequency zero point of the amplitude frequency response of the applied first RF excitation pulse (B1).

23. The apparatus of claim 22, wherein: The permanent magnet assembly includes a sample cavity.

24. The apparatus of claim 23, wherein: The housing is sized to be held in a tabletop stand, and the sample cavity is sized to receive a sample insert that holds the sample.

25. The apparatus of claim 23, wherein: The housing is sized to be held by a subject's hand, the sample is provided by a finger or thumb of the subject, and the sample cavity is sized to receive the subject's finger.

26. Apparatus according to any one of claims 23 to 25, wherein: The apparatus includes a sliding door to selectively allow access to the sample chamber during measurement.

27. Apparatus according to any one of claims 23 to 25, wherein: The device includes a touch sensor that can be touched to activate the device.

28. Apparatus according to any one of claims 23 to 25, wherein: The device includes a display for providing a user interface and displaying measurement results.

29. Apparatus according to any one of claims 23 to 25, wherein: The apparatus comprises a control unit for applying a magnetic excitation pulse to the sample and measuring a free induction decay signal generated by an ensemble of hydrogen protons within the sample; and a communication module for allowing the control unit to communicate with a remote device and receive control data and measurement results.

30. The apparatus of claim 22, wherein: The apparatus comprises a portable spectrometer (124), the portable spectrometer comprising: a radio frequency (RF) source configured to generate a pulsed RF signal having an in-phase component and a quadrature component; a transmitting unit configured to receive the pulsed RF signal and generate a pulsed RF excitation pulse (B1), the transmitting unit comprising: a first transmission path having a first transmit bandpass filter having at least one first transmit inductive coil, the at least one first transmit inductive coil receiving an in-phase component of the RF signal and generating an in-phase component of the RF excitation pulse; and a second transmission path having a second transmit bandpass filter having at least one second transmit inductive coil, the at least one second transmit inductive coil receiving the quadrature components of the RF signal and generating the quadrature components of the RF excitation pulses; a receiving unit configured to receive a resonance signal generated by the sample in response to the pulsed RF excitation pulse (B1), the receiving unit comprising: a first receiving path having a first receiving bandpass filter having at least one first receiving inductor coil receiving an in-phase component of the resonant signal; and a second receiving path having a second receiving bandpass filter having at least one second receiving inductor coil receiving a quadrature component of the resonant signal; and The processor unit is coupled to the RF source, the transmitting unit and the receiving unit, and the processor unit is configured to: control the operation of the portable spectrometer by sending a control signal to the RF source to generate the pulsed RF signal and sending the pulsed RF signal to the transmitting unit to generate the pulsed RF excitation pulse (B1), and the processor unit is configured to: receive the in-phase component and the orthogonal component of the resonance signal from the receiving unit.

31. The apparatus of claim 30, wherein at least one first transmit inductor coil of the first transmit bandpass filter, at least one second transmit inductor coil of the second transmit bandpass filter, at least one first receive inductor coil of the first receive bandpass filter, and at least one second receive inductor coil of the second receive bandpass filter are volume coils.

32. The apparatus according to claim 30, wherein each of the at least one first transmit inductance coil of the first transmit bandpass filter, the at least one second transmit inductance coil of the second transmit bandpass filter, the at least one first receive inductance coil of the first receive bandpass filter, and the at least one second receive inductance coil of the second receive bandpass filter is a tilted cosine coil coaxially positioned relative to each other and relative to a common axis, and wherein the at least one first transmit inductance coil of the first transmit bandpass filter and the at least one second transmit inductance coil of the second transmit bandpass filter are configured to generate an RF excitation pulse (B1) that is at least one of circular polarization and elliptical polarization.

33. The apparatus of claim 30, wherein: The resonance signal is a free induction decay (FID) signal generated by the decay of magnetization in the transverse plane, and the receiving unit uses the sum and difference of the FID signals received by the two tilted cosine coils to distinguish the projections of the magnetization on the two transverse spatial axes.

34. The apparatus of claim 30, wherein: The first receive bandpass filter and the second receive bandpass filter are configured to pass a second frequency passband range including at least one isotope Larmor frequency.

35. The apparatus of claim 32, wherein: The first receive bandpass filter and the second receive bandpass filter are configured to pass a second frequency passband range including at least one isotope Larmor frequency, wherein the at least one isotope Larmor frequency includes fluorine ( 19 F) and hydrogen ( 1 H) Larmor frequency.

36. The apparatus of claim 22, wherein: The permanent magnet assembly comprises an axisymmetric permanent magnet assembly rotating around a symmetry axis z, wherein the permanent magnet is configured to generate a uniform static magnetic field (B0) in a near space across the hole, wherein the permanent magnet assembly comprises: a top disk-conical magnet segment stacked above the aperture in the +z direction and having a magnetization in the +z direction; a bottom disk-conical magnet segment stacked below the aperture in the -z direction and having a magnetization in the +z direction; a central annular magnet segment positioned radially outward (r) from the aperture in the +r direction and having a magnetization in the -z direction, wherein the central annular magnet at least partially surrounds the aperture while maintaining an access opening to the aperture unobstructed; a top annular magnet segment stacked above the center annular segment in the +z direction and disposed radially outward from the top disk-conical magnet segment in the +r direction, wherein the top annular magnet segment has a magnetization in the +r direction; and a bottom annular magnet segment stacked below the center annular segment in the -z direction and arranged radially outward from the bottom disc-conical magnet segment in the +r direction, wherein the top annular magnet segment has a magnetization in the -r direction, wherein the superposition of the magnetic fields generated by each magnet segment generates a closely-spaced uniform static magnetic field (B0) across the bore along the axis of symmetry; and A shimming device is configured to provide spatial magnetic field correction to a near-space uniform static magnetic field.

37. The apparatus of claim 36, wherein: The permanent magnet assembly includes neodymium alloy (NdFeB) permanent magnet segments configured such that the superposition of magnetic fields generated by each neodymium alloy (NdFeB) permanent magnet segment provides a close-space uniform static magnetic field (B0) greater than 1.5 Tesla generated across the hole.

38. The apparatus of claim 36, wherein: The permanent magnet assembly includes neodymium alloy (NdFeB) permanent magnet segments, which are configured so that the superposition of the magnetic fields generated by each neodymium alloy (NdFeB) permanent magnet segment provides a near-space uniform static magnetic field (B0) generated across the hole of between 1.5 Tesla and 2 Tesla, in an arrangement weighing 10 kg and having a height and diameter of 10 cm each.

39. Apparatus according to any one of claims 36 to 38, wherein: The permanent magnet assembly includes neodymium alloy (NdFeB) permanent magnet segments configured so that the superposition of magnetic fields generated by each neodymium alloy (NdFeB) permanent magnet segment provides a near-space uniform static magnetic field (B0) generated across the hole with a field uniformity of less than 0.01 ppm.

40. The apparatus of claim 39, wherein: The permanent magnet assembly includes neodymium alloy (NdFeB) permanent magnet segments, which are configured so that the superposition of the magnetic fields generated by each neodymium alloy (NdFeB) permanent magnet segment provides a near-spaced uniform static magnetic field (B0) generated across the hole. Due to the correction provided by the field shim device, the near-spaced uniform static magnetic field (B0) generated across the hole has a field uniformity of substantially 0.01 ppm.

41. Apparatus according to any one of claims 36 to 38, wherein: The permanent magnet assembly includes neodymium alloy (NdFeB) permanent magnet segments configured such that the superposition of magnetic fields generated by each neodymium alloy (NdFeB) permanent magnet segment provides a three-dimensional confinement of 5 Gauss field lines.

42. The apparatus of claim 36, wherein: Each magnet segment of the permanent magnet assembly is formed of a hard permanent magnet alloy that allows for superposition of magnetic fields generated by each magnet segment in the permanent magnet assembly.

43. The apparatus of claim 42, wherein: The hard permanent magnet alloy is neodymium (NdFeB).

44. The apparatus of claim 43, wherein: The center ring magnet segment is formed of N40 grade NdFeB to counteract the high reverse coercive field, and the top and bottom ring magnet segments are formed of N40 grade NdFeB to provide optimal field confinement.

45. Apparatus according to any one of claims 36 to 38, wherein: The inner surface of the central annular magnet segment facing the hole includes a correction magnet segment that is curved radially inward and is configured to improve the uniformity of the close-space uniform static magnetic field (B0) generated across the hole, and the correction magnet segment is a pole piece formed of high permeability mild steel.

46. ​​The apparatus of claim 22, wherein: The apparatus comprises a shim accessory configured to provide spatial magnetic field correction for a static magnetic field having a nearly spatially uniform profile, wherein the shim accessory comprises a first set and a second set of concentric current-carrying conductors, wherein each concentric current-carrying conductor in the first set of concentric current-carrying conductors is spaced apart and arranged opposite to a corresponding concentric current-carrying conductor in the second set of current-carrying conductors along an axis of the static magnetic field to form a plurality of current-carrying conductor pairs, wherein each current-carrying conductor generates a compensating magnetic field to correct the nearly spatially uniform profile of the static magnetic field.

47. The apparatus of claim 46, each of the plurality of current carrying conductors being driven by a respective DC current.

48. The apparatus of claim 46, wherein: Each of the first and second sets of current carrying conductors includes a plurality of nested current carrying conductors, each current carrying conductor having a center point that is co-linear with an axis of the static magnetic field (B0).

49. The apparatus of claim 46, wherein: The magnetic field generated by each current carrying conductor is represented using a polynomial expansion of spherical harmonics, and each current carrying conductor generates even and odd polynomial terms, and current carrying conductors with different radii generate linearly independent polynomial coefficient vectors that allow any desired compensation field with azimuthal symmetry to be generated within the volume of the hole of the axisymmetric permanent magnet assembly, wherein the axisymmetric permanent magnet assembly generates the static magnetic field in the hole.

50. The apparatus of claim 46, wherein: The current carrying conductors in each of the first and second sets are offset from each other, and the offset is in a range between 10 and 45 degrees offset, and the shimming arrangement is configured to produce an azimuthally varying spatial magnetic field correction for a static magnetic field.

51. A compact, portable NMR device comprising a control unit configured to perform the method according to any one of claims 1 to 21.

52. A method for in vivo and non-invasive quantification, or in vitro and non-invasive quantification of the concentration of small molecule metabolites in a sample using a portable nuclear magnetic resonance (NMR) device, the method comprising: selecting a subset of high-resolution resonance features from a plurality of resonance features associated with the metabolite; causing magnetization of the sample by applying a uniform static magnetic field (B0) using a magnet assembly; Suppressing a water signal generated by magnetization of water in the sample, the sample comprising a glucose hydrogen proton ensemble, wherein suppressing the water signal comprises: applying a first radio frequency (RF) excitation pulse (B1) to rotate water magnetization into the transverse plane, wherein the first RF excitation pulse (B1) has an amplitude frequency response that reduces response to one or more chemical shift resonance frequencies associated with one or more of the glucose hydrogen proton ensembles located in the sample, the one or more of the glucose hydrogen proton ensembles having a chemical shift that coincides with a frequency zero of the amplitude frequency response of the applied first RF excitation pulse (B1), wherein the first RF excitation pulse (B1) is applied over a frequency range that includes at least one Larmor frequency associated with a resonance feature in the subset of the high resolution resonance features; applying a dephasing pulse gradient to the sample to dephase the coherence of the spin precession of water in the sample after applying the first RF excitation pulse (B1), and wherein the first RF excitation pulse (B1) and the dephasing pulse gradient suppress a water signal generated by the magnetization of the water in the sample; detecting a free induction delay (FID) signal generated by the sample; converting the FID signal into a magnetic resonance spectrum; and The concentration of the metabolite in the sample is determined based on one or more resonance peaks in the magnetic resonance spectrum, the one or more resonance peaks having a chemical shift coinciding with a frequency zero point of the amplitude frequency response of the applied first RF excitation pulse (B1).

Images