A gradient-free array magnetic particle imaging device

Abstract

A gradient-free array magnetic particle imaging device uses a miniaturized magnetic sensor array as a detection unit and realizes spatial positioning by utilizing the different relative spatial positions of the sensors relative to the scanned object. The magnetic sensors adopt a positive and negative symmetrical series structure to greatly reduce the interference of excitation noise on the detection signal. A multi-layer coil nesting mode can also be adopted to further reduce excitation interference and prevent magnetic sensor saturation. The system adopts an open imaging structure and has a larger open imaging space. Since the tracer is directly positioned by an array magnetic sensor, the magnetic particle imaging device is freed from the heavy dependence on high-intensity gradient magnetic fields, and is very convenient for setting and expanding any imaging interval, which has important technical advancement value for realizing clinical continuous dynamic imaging.

Description

Technical Field

[0001] The invention belongs to the technical field of medical detection devices, and in particular relates to an array-type magnetic particle imaging device that does not require a gradient positioning magnetic field. Background Art

[0002] Magnetic particle imaging is a new imaging method that uses superparamagnetic nanoparticles as tracers. It uses the principle that superparamagnetic particles produce nonlinear magnetic response in a high-frequency AC field to dynamically image the concentration and position of the tracer ingested into the human body. Its advantages are fast imaging speed, extremely high human safety, high sensitivity and spatial resolution, and it is suitable for coupling with a variety of targeted proteins or polymer structures, and has important scientific and clinical application value.

[0003] Existing MPI imaging equipment and systems generally require high-intensity and high-linearity gradient magnetic fields to spatially locate tracers. The basic principle is that the tracer reaches a magnetization saturation state under a high-intensity DC gradient field and no longer generates a magnetization signal under an AC excitation field, while the tracer at the zero point of the gradient field (generally referred to as the zero magnetic field point or zero magnetic field line) can still generate signals normally, thereby achieving the effect of spatial positioning scanning. The intensity of the DC gradient field directly determines the spatial resolution of MPI imaging. Usually, to obtain an imaging resolution of 1mm, the DC gradient field needs to reach 4-6T / m or more. In order to maintain a pure nonlinear magnetization response, MPI usually cannot use an iron core to enhance the gradient magnetic field. For MPI equipment with the volume of a human body, this intensity of gradient magnetic field is extremely difficult to achieve due to factors such as coil power, especially in the imaging method based on the Lissajous curve, where the AC excitation field and the static gradient field are coupled to form a magnetic field scanning trajectory. Under the general industry standard that the AC excitation field cannot be greater than 6mT (conditions that do not produce tissue heating and peripheral nerve stimulation), the size of the scanning space is usually limited to the size of small animals. Even if the Lisa-like AC / DC coupled scanning trajectory is not used, the large gradient field reaching human size will locally reach more than 3T, which does not meet the 3T static magnetic field intensity limit set by the current reference MRI technology in the field of human safety. Therefore, it is necessary to improve the existing magnetic particle imaging equipment in response to the needs of large-scale MPI imaging. Summary of the invention

[0004] In view of the above problems, the present invention proposes an array-type magnetic particle imaging device that does not require a gradient magnetic field. The device adopts a miniaturized magnetic sensor array as a detection unit, and realizes spatial positioning by utilizing the different relative spatial positions of the sensors relative to the scanned object. The magnetic sensors adopt a positive and negative symmetrical series structure to greatly reduce the interference of excitation noise on the detection signal. A multi-layer coil nesting mode can also be adopted to further reduce excitation interference and prevent magnetic sensor saturation. The system adopts an open imaging structure, has a larger open imaging space, and is also convenient for expanded use, which has important technical advancement value for realizing clinical continuous dynamic imaging, etc.

[0005] The technical solution adopted by the present invention is: the array-type magnetic particle imaging device with no gradient field includes a frame, a Helmholtz AC excitation coil group is arranged on the frame, and a uniformly arranged magnetic sensor array is symmetrically arranged for each Helmholtz excitation coil group; the magnetic sensor can use but is not limited to magnetic sensor devices such as magnetoresistive sensors (including MR, GMR, TMR, AMR, etc.) with high sensitivity, fluxgate sensors, atomic magnetometers, etc.; the open scanning field is set as a three-dimensional space directly in the Helmholtz coil group, and a system matrix is ​​formed by pre-scanning imaging results of superparamagnetic tracers of known concentrations at various points in space. When performing spatial imaging on unknown objects to be detected, the voltage distribution map can be directly output by using the positions of each sensor in the array sensor, or a concentration distribution map with higher resolution can be formed after reconstruction by the system matrix.

[0006] Preferably, the physical size of the magnetic sensor partially determines the spatial resolution of the magnetic particle imaging system. In particular, for the voltage distribution diagram directly output, using a magnetic sensor with a smaller size or a thin film magnetic sensor can achieve better spatial resolution.

[0007] Furthermore, the Helmholtz coil assembly includes a coil frame and an excitation coil wound on the coil frame. A high-frequency excitation current (frequency not higher than 20kHz) is passed through the excitation coil to generate a uniform excitation field between the two Helmholtz coils as the imaging area, and the relative direction of the two Helmholtz coils is defined as the Z axis, and the directions perpendicular to the Z axis are the X and Y axes. A power supply with an internal signal source is used, or an excitation signal is generated by a signal generator and amplified by a power amplifier to output the excitation current to the excitation coil. To ensure the safety of the magnetic field, the magnetic field strength in the imaging area is not higher than 6mT.

[0008] Preferably, the Helmholtz coil group is connected with the capacitor to form a series resonant circuit, and the resonance point of the series resonant circuit is ω 2 = LC ,in ω is the excitation signal angular frequency, L is the inductance of the Helmholtz coil, Cis the series capacitance value. Based on the measured coil inductance and the given signal angular frequency, the series capacitance value can be calculated. The series resonant circuit can reduce the AC impedance at the resonance point and apply a sufficiently strong excitation current to the Helmholtz coil.

[0009] Preferably, the Helmholtz coil assembly is wound with multiple strands of Litz wire. Since the excitation coil is an air-core coil, eddy current loss is the main loss form at the selected excitation frequency, and the alternately wound multiple strands of Litz wire can effectively reduce the eddy current loss between wires and the increase in equivalent DC resistance caused by the skin effect.

[0010] Preferably, since the excitation current flows through the coil to generate a magnetic field, the coil is a nonlinear load, and a more stable AC current output can be obtained by using a bipolar linear power supply or a power amplifier.

[0011] Furthermore, two groups of magnetic sensor arrays are arranged at the center of each coil of the Helmholtz coil group. The magnetic lines of force generated by the coils pass perpendicularly through the effective detection direction of the sensor. The two groups of sensor arrays are arranged oppositely with the X, Y plane located at the midpoint of the Z-axis direction of the coil as the axis. The output ends of the two sensors at relative positions are connected in series, that is, for the magnetic field strength H e , two reverse sensor input voltages connected in series V 1, V 2 are V 1 = H e × R and V 2 = - H e × R ,in R It is the ratio of the sensor input voltage to the input magnetic field strength, which is an inherent property of the sensor. When there is no detected signal, the series sensor group can reduce the interference of the excitation magnetic field, that is, V 1+ V 2≈0. All series sensor groups are labeled with array numbers, labeled as ( m , n , p ),in m is the row number, n is the column number, p Represents the Helmholtz excitation coil where the sensor array is located. For a set of Helmholtz excitation coils, define p = 1 or 2.

[0012] Preferably, due to factors such as uneven coil winding or magnetic sensor performance error, V 1+ V2 It is inevitable that it is difficult to completely zero the sensor in the fixed position. Therefore, a balance adjustment structure such as a thread is set between each group of series sensors so that the position of the two series sensors can be fine-tuned in the Z-axis direction relative to the Helmholtz coil to make V 1+ V 2 is as close to zero as possible to minimize the excitation noise interference.

[0013] Preferably, taking MR magnetic sensors as an example, MR sensors with high sensitivity can usually reach saturation in a lower background magnetic field strength, that is, they have reached a saturation state before the series effect takes effect, and further signal measurement cannot be performed. For this reason, an elimination coil wound in the opposite direction to the Helmholtz coil is set on the X, Y plane of the midpoint of each Helmholtz coil in the Z-axis direction, and the size of the elimination coil is much smaller than the external Helmholtz coil. The magnetic field frequency and phase generated by the elimination coil are the same as the excitation field, and the magnetic field intensity is equal to the magnetic field intensity generated by the Helmholtz at the array sensor position, but in the opposite direction, so as to offset the background interference of the excitation magnetic field itself at the array sensor position; and at a position far away from the Helmholtz and the elimination coil in the Z-axis direction, the excitation field intensity generated by the Helmholtz is much greater than the elimination coil, and the area within the uniform magnetic field is the actual imaging space.

[0014] Furthermore, when a structure with an external Helmholtz coil and an internal cancellation coil is used, each coil is powered by a power supply or a signal generator + power amplifier combination, and the frequency and phase of each power supply / power amplifier are the same.

[0015] Preferably, a group of power supplies / power amplifiers are used to power each coil simultaneously. In this case, a series resistor needs to be added according to the impedance of each coil so that when the power supply / power amplifier outputs a single voltage / current, a preset excitation current intensity is obtained in each coil.

[0016] Preferably, a signal generator is used to provide a signal source for two power amplifiers at the same time, and different excitation currents are provided to the Helmholtz excitation coil and the internal cancellation coil by setting different gain multiples for the power amplifiers. This method can better ensure that the currents in the Helmholtz coil and the internal cancellation coil are of the same frequency and phase.

[0017] Preferably, a variable resistor is connected in series in the elimination coil circuit, and the excitation current in the elimination coil is adjusted by fine-tuning the variable resistor to obtain the best excitation background noise elimination effect.

[0018] Preferably, since the excitation current flows through the excitation coil and the elimination coil to generate a magnetic field, the coil is a nonlinear load, and a more stable AC current output can be obtained by using a bipolar linear power supply or a power amplifier.

[0019] Furthermore, when the sample to be tested is placed in the open imaging area, each group of series sensors in the array sensor can obtain the magnetic induction intensity conversion voltage value corresponding to its position. The array sensor is connected to a multi-channel signal amplifier, which should have at least m × n × 2 input channels and output channels. The amplified signal is connected to the signal acquisition device. The signal acquisition device should have at least m × n × 2 signal input ports. The signal acquisition device inputs the acquisition signal to the host computer. Due to the large interference of the fundamental wave signal, the harmonic signal is used as the detection signal in the magnetic particle imaging system. The host computer performs Fourier transform on the acquired signal to remove the fundamental wave component and obtain each harmonic component.

[0020] Preferably, a bandpass filter, a lock-in amplifier or other circuit or device may be used to directly remove the fundamental signal and obtain each harmonic component;

[0021] Furthermore, the host computer reconstructs the image of the detection signal after removing the fundamental wave component through the image reconstruction method as described above. When the image resolution requirement is high and the reconstruction calculation amount is large, GPU computing or parallel computing is used to increase the image reconstruction speed, and the reconstructed image is displayed in the host computer.

[0022] Preferably, when the real-time imaging is required to be high, the host computer can directly display two voltage distribution diagrams corresponding to each sensor position while performing the next step of image reconstruction. The two voltage distribution diagrams represent the voltage signals measured by the sensor arrays located at the two Helmholtz coil positions. Suppose the scanning area size is x × y × z , because the voltage distribution map is not reconstructed, the distribution map is a two-dimensional image, and its size is x × y , the resolution is ( x / m )and( x / n ).

[0023] Furthermore, in order to achieve higher scanning resolution and obtain a three-dimensional scanning image, the voltage measurement results are reconstructed to obtain a more accurate magnetic particle concentration distribution map. V , which is related to the magnetic particle concentration distribution c Previously, the following relationship existed:

[0024] V = Ac

[0025] In the formula A is the system matrix.V Each voltage field point v xyz , which is related to the concentration area point c ijk The correlation can be expressed as:

[0026] v (x,y,z) =∑ (m,n,p) ∑ (i,j,k) a ((x,y,z),(i,j,k)) c (i,j,k)

[0027] In the formula a ((x,y,z),(i,j,k)) represent v (x,y,z) and c (i,j,k) The relevance of x is the X-direction coordinate of the voltage field, y is the Y-direction coordinate of the voltage field, z is the Z-direction coordinate of the voltage field, i is the X-direction coordinate of the concentration field, j is the Y-direction coordinate of the concentration field, k is the Z-direction coordinate of the concentration field, m is the row where the magnetic sensor is located, n is the column where the magnetic sensor is located, p Number the Helmholtz coils, p = 0 or 1. In the system function A When the concentration is known, the concentration can be solved by mathematical algorithms such as least squares method and singular value decomposition. c , whose image resolution is the system function density as described below.

[0028] Furthermore, the system function is obtained by measuring the system function A Use a high concentration of tracer samples that are no larger than the final resolution S A , the scanning space (concentration imaging space) is divided into x × y × z (= i × j × k ) points, so that the sample S A By placing them at all spatial points in the scanning range and imaging them, we can obtain S AA graph of voltage measurements obtained from each sensor in the sensor array as it is scanned across the imaging space, representing a sample of known concentration and position. S A , the voltage value measured by each sensor is S A The correlation of a ((x,y,z),(i,j,k)) value.

[0029] Preferably, x × y × z Each point in the space is measured to obtain the system function A The point spread function PSF (Point Spread Function) can be formed by using the measurement results obtained at the middle of the scanning area with a very high concentration point, and the center point of the PSF can be moved at each coordinate point in the measurement space to obtain the system function A The proposed method has an approximate solution and a faster measurement speed.

[0030] Further, yes a ((x,y,z),(i,j,k)) Perform matrix synthesis to obtain system functions for calculation A Matrix, starting from the (1, 1, 1)th sensor to the ( m , n , 2) The full scan space corresponding to each sensor x × y × z of a ((x,y,z),(i,j,k)) by x × y × z OK, x × y × z Columns form a matrix A 1, A 2… A m×n×2 ,Will A 1, A 2… A m×n×2 Arrange them vertically in series to get the system function A Matrix, the matrix size is x × y × z × x × y × z × m × n × 2.

[0031] Preferably, when the number of sensors is large and the spatial resolution requirement is high, the system matrix A The image is extremely large and takes a long time to reconstruct. The tomographic reconstruction method can be used to divide the Z direction image into 1 - z The fault plane corresponding to each sensor x × y of a ((x,y),(i,j)) by x × y OK, x × y Columns form a matrix A 1,z1 , A 1,z2 … A 1,zz ,Will A 1,z1 , A 1,z2 … A 1,zz Arrange in parallel and in series to obtain the system matrix of each sensor A 1, A 2… A m×n×2 ,Will A 1, A 2… A m×n×2 Arrange them vertically in series to get the system function A Matrix, the matrix size is x × y × z × x × y × m × n × 2.

[0032] In the array-type magnetic particle imaging device without gradient field, biofunctionalized superparamagnetic nanoparticles are used as tracers. The biofunctionalized superparamagnetic particle tracer is a composite nanomaterial, the core of which is a Fe2O3 or Fe3O4 magnetic core of several nanometers to tens of nanometers, and the outside of which is coated with a polymer coating to provide functions such as hydrophilicity / hydrophobicity, anti-aggregation surface potential and various functional groups. Superparamagnetic particles can be combined with a variety of biological molecules, such as proteins, enzymes, nucleic acids, etc., through the functional groups of the polymer coating. For example, the functional groups modified with carboxyl groups can form amide bonds with the free amino groups on the surface of antibodies, so that the antibodies are stably connected to the surface of the magnetic sphere to form biofunctionalized superparamagnetic nanoparticles.

[0033] The imaging steps of the magnetic particle imaging system are as follows:

[0034] Step 1: Pass an AC excitation current into the Helmholtz coil or into the Helmholtz coil and the elimination coil at the same time to form a uniform spatial excitation magnetic field in the open scanning area;

[0035] Step 2: Place a sample containing a known high concentration of superparamagnetic particle tracer in the scanning area, and the sample size should not exceed the spatial resolution requirement. Each sensor in the magnetic sensor array will obtain the magnetization signal of the sample. The magnetization signal is converted into a voltage signal by the sensor, amplified by the signal amplifier, and output to the host computer through a multi-channel signal acquisition device. The host computer performs Fourier transform on the input signal, removes the fundamental component of the signal and obtains each harmonic component, which is used as the original effective voltage signal collected by each sensor;

[0036] Preferably, in step 2, a bandpass filter, a lock-in amplifier or other circuit or device may be used to directly remove the fundamental signal and obtain each harmonic component;

[0037] Step 3: Move the known sample to all positions in the scanning area, measure the signal voltage value obtained by each magnetic sensor at each spatial position, and form a system function A When the system itself does not change, the system function A No changes occur, no need to repeat measurements;

[0038] Preferably, a point spread function (PSF) is used to place a sample with a very high concentration at the center of the scanning area, and a system function is formed by moving the center point of the PSF in the scanning area. A An approximate solution to reduce the system function measurement time;

[0039] Step 4: Place the actual object under test into an open imaging space, perform signal detection on it, and obtain a voltage distribution diagram through the position of each sensor in the array sensor and the measured voltage value;

[0040] Step 5: Use each sensor to measure the voltage value and system function A , the voltage concentration distribution map is obtained by matrix calculation methods such as least squares method and singular value decomposition c and display it.

[0041] Beneficial effects of the present invention: Compared with traditional magnetic particle imaging equipment and systems, since the present invention directly uses array-type magnetic sensors to locate tracers, the magnetic particle imaging equipment is freed from the heavy reliance on high-intensity gradient magnetic fields, and is very convenient for setting and expanding any imaging interval, which plays an important role in the technical improvement of the current MPI technology from animal level to human use. At the same time, the present invention reduces the excitation background interference when using weak magnetic sensors by arranging small weak magnetic sensors in series in forward and reverse directions, and can further reduce the fundamental background noise interference problem by reverse elimination coils, making MPI signal detection by array-type weak magnetic sensors more practical. In addition, the magnetic particle imaging image reconstruction algorithm for array-type weak magnetic sensors proposed in the present invention can further improve the resolution of magnetic particle imaging while achieving three-dimensional image reconstruction.

[0042] As a medical imaging system, compared with other in vivo imaging technologies (such as MRI or CT, etc.), the present invention has extremely high dynamic imaging speed and sensitivity, and has the characteristics of open imaging space, no ionizing radiation, high human safety of superparamagnetic particle tracers, and easy coupling of various targeting structures to achieve targeted imaging. It is a good new technology support and supplement for the preclinical and clinical applications of medical imaging. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a structural schematic diagram of the array-type magnetic particle imaging device without a gradient field of the present invention.

[0044] Figure 2 It is a system flow chart of an array-type magnetic particle imaging device without a gradient field.

[0045] Figure 3 It is a schematic diagram of the system function measurement principle in the present invention.

[0046] Figure 4 This is the voltage distribution diagram of the “big” shaped sample and the reconstructed result.

[0047] Figure 5 It is the simulation calculation result of the voltage distribution in the detection area formed by the Helmholtz excitation coil and the elimination coil.

[0048] Figure 6 is the system matrix used for 3D image reconstruction.

[0049] Figure 7 is a simplified system matrix for 3D tomography.

[0050] Explanation of the serial numbers in the figure: 1 Helmholtz line group, 2 elimination coil group, 3 magnetic sensor array, 4 scanning area, 5 elimination coil resonance, 6 Helmholtz excitation coil resonance, 7 elimination coil power amplifier (or power supply of built-in signal source), 8 excitation coil power amplifier (or power supply of built-in signal source), 9 multi-channel signal amplifier, 10 multi-channel signal acquisition device, 11 host computer, 12 signal generator, 13 magnetic lines of force generated by the excitation magnetic field. DETAILED DESCRIPTION

[0051] according to Figures 1 - 4 The specific structure of the present invention is described in detail. The array imaging device includes a Helmholtz AC excitation coil group 1, an elimination coil 2, a magnetic sensor array 3, an elimination coil resonance 5, a Helmholtz excitation coil resonance 6, an elimination coil power supply 7, an excitation coil power supply 8, a multi-channel signal amplifier 9, a multi-channel signal acquisition device 10, a host computer 11 and other devices.

[0052] An open scanning area 4 is formed in the middle of the Helmholtz coil group 1, and the open scanning area 4 is set as a three-dimensional space between the Helmholtz coil groups 1. A uniformly arranged magnetic sensor array 3 is symmetrically arranged for each Helmholtz excitation coil group 1, and the magnetic sensor array 3 is located inside the elimination coil group 2. The magnetic sensor can be used but not limited to magnetic sensor devices such as magnetoresistive sensors (including MR, GMR, TMR, AMR, etc.) with high sensitivity, fluxgate sensors, and atomic magnetometers.

[0053] The Helmholtz AC excitation coil assembly 1 can be wound with multiple strands of Litz wire. Since the Helmholtz AC excitation coil assembly 1 is a hollow coil, eddy current loss is the main loss form at the selected excitation frequency, and the alternately wound multiple strands of Litz wire can effectively reduce the eddy current loss between wires and the increase in equivalent DC resistance caused by the skin effect.

[0054] A high frequency excitation signal is generated by the signal generator 12, which is amplified by the power amplifier 8 and then input into the Helmholtz coil assembly 1 to generate an AC excitation magnetic field, or a high power supply with an internal signal generator is used to directly input an AC excitation current.

[0055] Two groups of magnetic sensor arrays 3 are arranged at the center of each coil of the Helmholtz excitation coil group 1. The magnetic lines of force generated by the coil 1 vertically pass through the magnetic sensor array 3 to effectively detect the direction. The two groups of sensor arrays 3 are arranged oppositely with the X, Y plane located at the midpoint of the Z axis direction of the coil 1 as the axis. The output ends of the two sensors 3 at relative positions are connected in series, that is, for the magnetic field strength H e , two reverse sensor input voltages connected in series V 1, V 2 are V 1 =H e × R and V 2 = - H e × R ,in R is the ratio of the input voltage of sensor 3 to the input magnetic field strength, which is an inherent property of sensor 3. When there is no detected signal, the series sensor group 3 can reduce the interference of the excitation magnetic field, that is, V 1 + V 2≈ 0. All the series sensor groups 3 are labeled with array numbers, labeled as ( m, n, p ),in m is the row number, n is the column number, p Represents the Helmholtz AC excitation coil 1 where the sensor array 3 is located. For a group of Helmholtz excitation coils 1, define p = 1 or 2.

[0056] Due to factors such as uneven winding of coil 1 or performance error of magnetic sensor array 3, V 1 + V 2 It is inevitable that it is difficult to completely zero the sensor in the fixed position. Therefore, a balance adjustment structure such as a thread is set between each group of series sensors 3, so that the position of the two series sensors can be fine-tuned in the Z-axis direction relative to the Helmholtz AC excitation coil 1, so that V 1 + V 2 is as close to zero as possible to minimize the excitation noise interference.

[0057] Preferably, taking MR magnetic sensors as an example, MR sensors with high sensitivity can usually reach saturation in a lower background magnetic field strength, that is, they have reached a saturation state before the series effect takes effect, and further signal measurement cannot be performed. For this reason, an elimination coil group 2 is set on the X, Y plane at the midpoint of each Helmholtz AC excitation coil 1 in the Z-axis direction, which is wound in the opposite direction to the Helmholtz AC excitation coil 1, and the size of the elimination coil group 2 is smaller than the external Helmholtz coil 1. The magnetic field frequency and phase generated by the elimination coil 2 are the same as the excitation field, and the magnetic field intensity is equal to the magnetic field intensity generated by the Helmholtz AC excitation coil group 1 at the position of the array sensor 3, but in the opposite direction, so as to offset the background interference of the excitation magnetic field itself at the position of the magnetic sensor array 3; and at a position far away from the Helmholtz AC excitation coil group 1 and the elimination coil group 2 in the Z-axis direction, the excitation field intensity generated by the Helmholtz AC excitation coil group 1 is much greater than that of the elimination coil group 2, and the area within which the magnetic field is uniform is the actual imaging space. A variable resistor is connected in series in the circuit of the elimination coil group 2, and the excitation current in the elimination coil group 2 is adjusted by fine-tuning the variable resistor to obtain the best excitation background noise elimination effect. This implementation method uses a method of setting an elimination coil to reduce the excitation background noise.

[0058] When using a structure with an external Helmholtz coil 1 and an internal elimination coil 2, each coil is powered by a power supply or a combination of a signal generator 12 + power amplifiers 7, 8, and each power supply / power amplifier has the same frequency and phase. A group of power supplies / power amplifiers can also be used to power each coil at the same time. In this case, a series resistor needs to be added according to the impedance of each coil so that when the power supply / power amplifier outputs a single voltage / current, each coil obtains a preset excitation current intensity.

[0059] Preferably, a signal generator 12 can be used to provide a signal source for the two power amplifiers 7 and 8 at the same time, and different excitation currents can be provided for the Helmholtz excitation coil 1 and the internal elimination coil 2 by setting different gain multiples for the power amplifiers 7 and 8. This method can better ensure that the currents in the Helmholtz coil 1 and the internal elimination coil 2 are of the same frequency and phase. This power supply method is adopted in this embodiment. At the same time, in order to ensure the safety of the magnetic field, the magnetic field strength in the scanning area is lower than 6mT, and the excitation frequency is not higher than 20kHz.

[0060] In this array imaging device, biofunctionalized superparamagnetic nanoparticles are used as tracers. The biofunctionalized superparamagnetic particle tracer is a composite nanomaterial, the core of which is a Fe2O3 or Fe3O4 magnetic core of several nanometers to tens of nanometers, and the outside of which is coated with a polymer coating to provide functions such as hydrophilicity / hydrophobicity, anti-aggregation surface potential and various functional groups. Superparamagnetic particles can be combined with a variety of biological molecules, such as proteins, enzymes, nucleic acids, etc., through the functional groups of the polymer coating. For example, the functional groups modified with carboxyl groups can form amide bonds with the free amino groups on the surface of antibodies, thereby stably connecting the antibodies to the surface of the magnetic spheres to form biofunctionalized superparamagnetic nanoparticles.

[0061] When the sample is placed in the open imaging area 4, each group of series sensors in the array sensor can obtain the magnetic induction intensity conversion voltage value corresponding to its position. The array sensor is connected to a multi-channel signal amplifier, which should have at least m × n × 2 input channels and output channels. The amplified signal is connected to the signal acquisition device. The signal acquisition device should have at least m × n × 2 signal input ports. The signal acquisition device inputs the acquisition signal to the host computer. Due to the large interference of the fundamental wave signal, the harmonic signal is used as the detection signal in the magnetic particle imaging system. The host computer performs Fourier transform on the acquired signal to remove the fundamental wave component and obtain each harmonic component.

[0062] After signal amplification and multi-channel acquisition, two voltage distribution maps corresponding to the sensor position can be directly obtained. The two distribution maps are the images obtained by the sensor array located in the direction of the two Helmholtz coils. Assume that the scanning area size is x × y × z , because the voltage distribution map is not reconstructed, the distribution map is a two-dimensional image, and its size is x × y , the resolution is ( x / m )and( x / n ).

[0063] In order to achieve higher scanning resolution and obtain three-dimensional scanning images, the voltage measurement results are reconstructed to obtain a more accurate magnetic particle concentration distribution map. V , which is related to the magnetic particle concentration distribution c Previously, the following relationship existed:

[0064] V = Ac

[0065] In the formula A is the system matrix. V Each voltage field point v xyz , which is related to the concentration area point c ijk The correlation can be expressed as:

[0066] v (x,y,z) =∑ (m,n,p) ∑ (i,j,k) a ((x,y,z),(i,j,k)) c (i,j,k)

[0067] In the formula a ((x,y,z),(i,j,k)) represent v (x,y,z) and c (i,j,k) The relevance of x is the X-direction coordinate of the voltage field, y is the Y-direction coordinate of the voltage field, z is the Z-direction coordinate of the voltage field, i is the X-direction coordinate of the concentration field, j is the Y-direction coordinate of the concentration field, k is the Z-direction coordinate of the concentration field, m is the row where the magnetic sensor is located, n is the column where the magnetic sensor is located, p Number the Helmholtz coils, p = 0 or 1. In the system function AWhen the concentration is known, the concentration can be solved by mathematical algorithms such as least squares method and singular value decomposition. c , whose image resolution is the system function density.

[0068] Obtaining the system function by measuring the system function A Use a high concentration of tracer samples that are no larger than the final resolution S A , the scanning space (concentration imaging space) is divided into x × y × z ( =i × j × k ) points, the number of divisions represents the system function density, that is, the spatial resolution after image reconstruction. S A By placing them at all spatial points in the scanning range and imaging them, we can obtain S A A graph of voltage measurements obtained from each sensor in the sensor array as it is scanned across the imaging space, representing a sample of known concentration and position. S A , the voltage value measured by each sensor is S A The correlation of a (( x, y, z ),( i, j, k ))value.

[0069] right x × y × z Each point in the space is measured to obtain the system function A The point spread function PSF (Point Spread Function) can be formed by using the measurement results obtained at the middle of the scanning area with a very high concentration point, and the center point of the PSF can be moved at each coordinate point in the measurement space to obtain the system function A The proposed method has an approximate solution and a faster measurement speed.

[0070] For a(( x, y, z ),( i, j, k )) Perform matrix synthesis to obtain the system function used for calculation A Matrix, starting from the (1,1, 1)th sensor to the ( m, n 2) The full scan space corresponding to each sensor x × y × z of a (( x, y, z ),( i, j, k ))by x × y × z OK, x × y × z Columns form a matrix A 1, A 2… Am×n×2 ,Will A 1, A 2… A m×n×2 Arrange them vertically in series to get the system function A Matrix, the matrix size is x × y × z × x × y × m × n × 2.

[0071] When there are many sensors and the spatial resolution requirement is high, the system matrix A The image is extremely large and the image reconstruction takes a long time. The tomographic reconstruction method can be used to divide the Z direction image into 1-z tomographic layers, and the x × y tomographic plane corresponding to each sensor is a (( x, y ),( i, j ))by x × y OK, x × y Columns form a matrix A 1, z 1, A 1, z 2… A 1, zz ,Will A 1, z 1, A 1, z 2… A 1, zz Arrange in parallel and in series to obtain the system matrix of each sensor A 1, A 2… A m×n×2 ,Will A 1, A 2… A m×n×2 Arrange them vertically in series to get the system function A Matrix, the matrix size is x × y × z × x × y × m × n × 2. This embodiment uses a tomographic reconstruction method to obtain the system function matrix A .

[0072] The voltage distribution measured by the sensor array is analyzed using the non-negative least squares method NNLS in the host computer. V Reconstruct the system function matrix obtained using the above-mentioned tomographic reconstruction method A , the tracer spatial concentration distribution map can be obtained and displayed. A Large, use GPU or parallel computing to speed up the calculation.

[0073] The imaging steps of the magnetic particle imaging device are as follows:

[0074] Step 1: A sinusoidal excitation current is introduced into the Helmholtz excitation coil group 1 to generate a highly uniform AC excitation magnetic field in the imaging space region 4 .

[0075] Step 2: An AC excitation current with the same frequency and phase as those of the Helmholtz coil group 1 and opposite direction is passed into the elimination coil group 2 inside the Helmholtz excitation coil group 1. The current in the reverse direction of the elimination coil group 2 is smaller than the current inside the Helmholtz excitation coil group 1, thereby offsetting the background interference of the excitation magnetic field itself at the position of the array sensor 3.

[0076] Step 3: Place a sample containing a known high concentration of superparamagnetic particle tracer in the scanning area 4, and the sample size is not greater than the spatial resolution requirement. Each sensor in the magnetic sensor array 3 will obtain the magnetization signal of the sample, which is converted into a voltage signal by the array sensor 3, amplified by the signal amplifier 9, and output to the host computer 11 through the multi-channel signal acquisition device 10. The host computer 11 performs Fourier transform on the input signal, removes the fundamental component of the signal and obtains each harmonic component, which is used as the original effective voltage signal collected by each sensor.

[0077] Step 4: Move the known sample to all positions in the scanning area 4, and measure the signal voltage value obtained by each magnetic sensor 3 at each spatial position of the superparamagnetic particle sample through the array sensor 3 to form a system function A When the system itself does not change, the system function A No changes, no need to repeat measurements.

[0078] Preferably, a point spread function (PSF) is used to place a sample with a very high concentration at the center of the scanning area, and a system function is formed by moving the center point of the PSF in the scanning area. A to reduce the system function measurement time.

[0079] Step 5: Place the actual object to be measured into the open imaging space 4, perform signal detection on it, and obtain a voltage distribution diagram through the position of each sensor in the array sensor 3 and the measured voltage value.

[0080] Step 6: Use each sensor to measure the voltage value and system function A , the voltage concentration distribution map is obtained by matrix calculation methods such as least squares method and singular value decomposition c and display it.

[0081] Example:

[0082] like Figure 1As shown, when the present invention works, the illustrated equipment is used, the Helmholtz coil group power supply 8 adopts the 4610 power amplifier of NF Company, the elimination coil group power supply 7 adopts the 4610 power amplifier of NF Company, the signal amplifier 9 adopts the SA-440F5 signal amplifier of NF Company, and the signal acquisition device 10 adopts the NI PCI-6534 data acquisition card.

[0083] In this embodiment, a group of Helmholtz coils with a diameter of 30 cm are used as excitation coils, with 270 turns, an excitation current intensity of 1A, and a magnetic field intensity of 2mT in the center of the Helmholtz coil; a group of coils with a diameter of 8 cm are used as elimination coils, with 50 turns and a current intensity of 0.2A. The frequencies of the excitation current and the elimination current are both 20kHz. The same signal generator is used to generate the signal source, and a BNC three-way interface is used to connect two power amplifiers to apply the same frequency and same phase excitation current to the coils. The excitation coil and the elimination coil jointly generate a uniform excitation field area near the center of the Helmholtz coil group 1, and its size is approximately 80mm×80mm×70mm. The simulation calculation results of the uniform excitation area are shown as follows. Figure 5 shown.

[0084] Since the high-frequency AC impedance of the Helmholtz coil group 1 and the elimination coil group 2 is relatively large, the series resonance 6 and the series resonance 5 are used to reduce the impedance of the Helmholtz coil group 1 and the elimination coil group 2. The series resonance 6 and the series resonance 5 resonate with the Helmholtz coil and the elimination coil at a frequency of 20kHz respectively. In addition, in order to avoid the increase of the equivalent DC resistance of the coil caused by the skin effect of the current and effectively prevent thermal noise interference, the Helmholtz coil group 1 and the elimination coil group 2 are wound with multiple strands of Litz wire.

[0085] A uniformly arranged magnetic sensor array 3 is symmetrically arranged near each Helmholtz excitation coil 1; the magnetic sensor can be used but not limited to a highly sensitive magnetoresistive sensor, and the superparamagnetic particle tracer in the sample is detected by the magnetic sensor array 3. The magnetic sensor array 3 used in this embodiment is composed of a multi-dimensional technology TMR9001 linear magnetic field sensor, the size of a single sensor is 6mm × 5mm × 1.5mm, and the magnetic sensors are arranged in a group in an 8 × 8 rectangle, with a 2mm interval between two magnetic sensors, and 4 groups of 8 × 8 magnetic sensors are arranged in a vertically symmetrical structure on the upper and lower sides of the Helmholtz coil group 1, respectively. The two magnetic sensors symmetrically connected in series with each other about the Helmholtz coil adopt a differential connection method to form 2 groups of 8 × 8 sensor arrays, and the array size is 62mm × 54mm, which is smaller than the XY plane size of the uniform magnetic field area, so the imaging area size is set to 62mm × 54mm × 70mm.

[0086] The tracer uses the screened Resovist superparamagnetic particles of Fuji Film Company, and the average size of the screened superparamagnetic core is about 21nm, and the hydrodynamic diameter is about 60nm. After placing the sample to be tested in the scanning area, each group of series sensors in the magnetic sensor array 3 can obtain the magnetic induction intensity corresponding to its position and convert it into a voltage value. After the signal amplifier 9 and the signal acquisition card 10, the data information is transmitted to the host computer 11, and two voltage distribution maps corresponding to the sensor position can be directly obtained. The two distribution maps are the images obtained by the sensor array located in the direction of the two Helmholtz coils. Because the size of the scanning area is 62mm × 54mm × 70mm, the voltage distribution map is not reconstructed, and the distribution map obtained is a two-dimensional image, which is 62mm × 54mm in size and has a resolution of (62 / 8 = 7.75mm) and (54 / 8 = 6.75mm). Figure 4 The left part shows the large font sample placed in a magnetic field. z Voltage distribution diagram of the single-sided array measurement in the plane = 0 (center point of the Helmholtz coil).

[0087] Furthermore, in order to achieve higher scanning resolution and obtain a three-dimensional scanning image, the voltage measurement result is reconstructed to obtain a more accurate magnetic particle concentration distribution map. c Previously, the following relationship existed:

[0088] V = Ac

[0089] In the formula A is the system matrix. V Each voltage field point v xyz , which is related to the concentration area point c ijk The correlation can be expressed as:

[0090] v( x, y, z )=∑( m, n, p )∑( i, j, k ) a(( x, y, z ),( i, j, k ))c( i, j, k )

[0091] In the formula, a(( x, y, z ),( i, j, k )) represents v( x, y, z ) and c( i, j, k ), x is the X-direction coordinate of the voltage field, y is the Y-direction coordinate of the voltage field, z is the Z-direction coordinate of the voltage field,i is the X-direction coordinate of the concentration field, j is the Y-direction coordinate of the concentration field, k is the Z-direction coordinate of the concentration field, m is the row where the magnetic sensor is located, n is the column where the magnetic sensor is located, p Number the Helmholtz coils, p = 0 or 1.

[0092] Furthermore, the system function is obtained by measuring the system function A Use a high concentration of tracer samples that are no larger than the final resolution S A (200 mg / mL), the scanning space (concentration imaging space) was segmented into 62×54×70 (= x × y × z = i × j × k ) points, that is, the system function density (concentration reconstruction resolution) is 1mm. S A By placing them at all spatial points in the scanning range and imaging them, we can obtain S A A graph of voltage measurements obtained from each sensor in the sensor array as it is scanned across the imaging space, representing a sample of known concentration and position. S A , the voltage value measured by each sensor is S A The correlation of can be obtained by a( (x, y, z ),( i, j, k )) value, the system function measurement method is as follows Figure 3 shown.

[0093] Preferably, x × y × z Each point in the space is measured to obtain the system function A It is time-consuming, so the measurement results obtained by using a very high concentration point in the middle of the scanning area form a point spread function (PSF), and the center point of the point spread function is assigned different position coordinates in the imaging area to obtain the system function. A The proposed method has an approximate solution and a faster measurement speed.

[0094] Further, yes a (( x, y, z ),( i, j, k )) Perform matrix synthesis to obtain the system function used for calculation A Matrix, starting from the (1, 1, 1)th sensor to the ( m , n , 2) The full scan space corresponding to each sensorx × y × z of a (( x, y, z ),( i, j, k ))by x × y × z OK, x × y × z Columns form a matrix A 1, A 2… A m×n×2 ,Will A 1, A 2… A m×n×2 Arrange them vertically in series to get the system function A Matrix, the matrix size is x × y × z × x × y × z × m × n × 2 , the specific matrix format is as follows Figure 6 shown.

[0095] Preferably, since the number of sensors used in the embodiment is 64 and the system matrix A The image is large and the image reconstruction takes a long time. Therefore, the tomographic reconstruction method is used to divide the Z direction image into 1-z tomographic planes, and the tomographic plane corresponding to each sensor is x × y of a (( x, y ),( i, j ))by x × y OK, x × y Columns form a matrix A 1, z 1, A 1, z 2… A 1, zz ,Will A 1, z 1, A 1, z 2… A 1, zz Arrange in parallel and in series to obtain the system matrix of each sensor A 1, A 2… A m×n×2 ,Will A 1, A 2… A m×n×2 Arrange them vertically in series to get the system function A Matrix, the matrix size is x × y × z × x × y × m × n × 2 , the specific matrix format is as follows Figure 7 shown.

[0096] The voltage measurement results are reconstructed using the non-negative least squares method NNLS. The calculation results are as follows: Figure 4 As shown on the right, the sample is placedz = 0 plane, the calculation results are 0 in other faults. Figure 4 Show only z = 0 plane concentration distribution map, the image resolution is greatly improved after reconstruction. The image pixel discontinuity is considered to be due to the large size of the sensor and the arrangement interval of about 2mm. A more miniaturized sensor or TMR film can be used to form a sensor array.

Claims

1. A non-gradient array magnetic particle imaging device, comprising an excitation coil (1), an array magnetic sensor (3), an open scanning area (4) and a host computer (11), characterized in that: The excitation coil (1) adopts a Helmholtz coil group structure, and the open scanning area (4) is located in the uniform field area in the middle of the Helmholtz coil group; two groups of array magnetic sensors (3) are arranged at the center of each coil of the Helmholtz coil group, and the magnetic lines of force generated by the Helmholtz coil group vertically pass through the effective detection direction of the array magnetic sensor (3); the two groups of array magnetic sensors (3) are arranged in opposite directions with the X, Y plane located at the midpoint of the Z-axis direction of the Helmholtz coil group as the axis, and the output ends of the two array magnetic sensors (3) in relative positions are connected in series to offset the interference of the excitation background noise; the array magnetic particle imaging device adopts superparamagnetic nanoparticles as tracers, and each sensor in the array magnetic sensor (3) collects the magnetization signal of the tracer in the signal scanning area, and converts it into a voltage value and outputs it to the host computer (11); the image reconstruction unit of the host computer (11) performs image reconstruction on the voltage value according to a pre-determined system function to obtain a tracer concentration distribution map; in the Helmholtz coil group and in the Helmholtz coil The core wire is coaxially wound with an elimination coil group (2); the excitation coil (1) and the elimination coil group (2) have the same frequency and phase; the magnetic field intensity generated by the excitation coil (1) at the position of the array magnetic sensor is the same as the magnetic field intensity generated by the elimination coil group (2) but in the opposite direction, so as to offset the excitation background noise of the array magnetic sensor (3) and prevent sensor saturation; near the center of the Helmholtz coil group, the magnetic field intensity generated by the excitation coil (1) is much greater than that of the elimination coil group (2), which serves as a scanning area for magnetic particle imaging; the array magnetic sensor (3) is located inside the elimination coil group (2).

2. The array-type magnetic particle imaging device without gradient field according to claim 1, characterized in that: The excitation coil (1) and the elimination coil group (2) reduce the excitation frequency AC impedance through a series resonant circuit (5, 6); the voltage signal output by the array magnetic sensor (3) is output to a multi-channel signal acquisition device (10) via a signal amplifier (9) and then to a host computer (11); the host computer (11) separates the fundamental wave component through Fourier transform and collects each harmonic component as an original effective voltage signal.

3. The array-type magnetic particle imaging device without gradient field according to claim 1, characterized in that: When a faster MPI scanning speed is required, the host computer can directly output or display a two-dimensional voltage cloud map directly measured by the array magnetic sensor, and the coordinates of the voltage cloud map correspond to the positions of the magnetic sensors in the array magnetic sensor.

4. The array-type magnetic particle imaging device without gradient field according to claim 1, characterized in that: For two magnetic sensors connected in series at corresponding positions in the oppositely arranged array magnetic sensors, their positions relative to the Helmholtz coil group can be finely adjusted for balance to compensate for the imbalance of positive and negative signals caused by coil winding or sensor errors.

5. The array-type magnetic particle imaging device without gradient field according to claim 1, characterized in that: Two groups of excitation power supplies with the same frequency and phase are used to output excitation current to the Helmholtz coil group and the elimination coil group (2) respectively; or a series resistor is used to change the total resistance of the circuit where the coils are located, and then a group of excitation power supplies is used to output excitation current to the Helmholtz coil group and the elimination coil group (2) at the same time.

6. The array-type magnetic particle imaging device without gradient field according to claim 5, characterized in that: A variable resistor connected in series with the elimination coil group (2) is used to fine-tune the current in the elimination coil group (2) so as to obtain a better excitation background noise elimination effect at the position of the array magnetic sensor (3) and a more uniform excitation magnetic field at a position far away from the array magnetic sensor (3).

Images