Method and system for magnetic particle acoustic imaging excited by ultra-short pulsed x-ray

Abstract

The embodiment of the application provides a kind of super-short pulse X-ray excited magnetic particle acoustic imaging method and system, the system includes: X-ray signal transmitting module, for responding to control instruction, control X-ray generator generates the X-ray pulse of preset parameter, so that the X-ray pulse acts on the magnetic particle mark region of pre-setting;Magnetic field modulation module is used to modulate the magnetic field of the magnetic particle mark region to make the X-ray pulse act on magnetic particle and produce original acoustic signal matched with magnetic field modulation;Acoustic detection module is used to convert the original acoustic signal into original data signal;Signal processing module is used to receive the original data signal, and the original data signal is processed to obtain imaging result.The super-short pulse X-ray excited magnetic particle acoustic imaging method and system of the application can meet long-term monitoring demand, effectively distinguish between viable cells and free magnetic particles, avoid misjudgment, and provide complete information for diagnosis and treatment.

Description

Technical Field

[0001] This application relates to the field of magnetic particle imaging, specifically to a method and system for magnetic particle acoustic imaging excited by ultrashort pulse X-rays. Background Technology

[0002] In the fields of regenerative medicine and tumor immunotherapy, cell tracing technology is a core tool for evaluating treatment efficacy and optimizing treatment plans. Stem cell transplantation has entered clinical research stages in the treatment of diseases such as Parkinson's disease (dopaminergic stem cells replace damaged neurons), spinal cord injury (neural stem cells repair damaged pathways), and stroke (mesenchymal stem cells promote nerve regeneration). Magnetic particle-labeled T-cell immunotherapy (such as CAR-T cell therapy) has also become an important direction for the treatment of hematological malignancies and solid tumors. However, existing cell tracing technologies cannot simultaneously meet the triple requirements of "high spatial resolution, high temporal resolution, and high sensitivity," thus hindering the development of precision medicine. Traditional cell tracing techniques are mainly divided into three categories, all of which have obvious drawbacks: First, optical imaging techniques (such as fluorescence imaging and bioluminescence imaging), although simple to operate, have limited tissue penetration depth (≤2cm), making it impossible to trace cells in deep tissues (such as the brain and spinal cord), and are easily affected by autofluorescence of biological tissues, resulting in low sensitivity (only ≥10³ cells can be detected); Second, MRI imaging techniques, although with deep penetration depth, have poor temporal resolution (a single scan requires 5-10 minutes), making it impossible to track the dynamic migration process of cells in real time, and the spatial resolution can only reach 1-2mm, making it difficult to identify a small number of transplanted cells (such as ≤10² stem cells).

[0003] In summary, the existing technology mainly has the following problems: 1. Difficulty in achieving multi-dimensional performance: Existing technologies cannot simultaneously meet the requirements of "deep penetration (≥15cm), high spatial resolution (≤250μm), high temporal resolution (≤100μs / frame), and high sensitivity (≤100 cells)". For example, although MRI has deep penetration, its resolution and temporal performance are poor, while ultrasound has high resolution but shallow penetration.

[0004] 2. Insufficient long-term tracking capability: fluorescent probes and other markers are easily degraded or photobleached, making it impossible to achieve long-term cell tracking for more than 7 days. However, stem cell transplantation therapy (such as spinal cord injury repair) requires monitoring cell survival and migration for several months, which current technology cannot meet.

[0005] 3. High radiation risk or excessive equipment complexity: Some X-ray imaging technologies (such as CT) have high radiation doses (single scan dose ≥5mSv), and long-term monitoring can easily pose health risks to patients; while synchrotron radiation X-ray equipment is bulky (occupying ≥50m²) and expensive (≥10 million yuan), making it impossible to widely use in clinical institutions.

[0006] 4. Cell state identification and missing key parameters: It can only detect the distribution of magnetic particles, but cannot distinguish between living cells and free magnetic particles, and cannot simultaneously obtain key parameters such as blood vessel distribution, making it difficult to comprehensively evaluate the treatment effect. Summary of the Invention

[0007] This application aims to provide a method and system for magnetic particle acoustic imaging excited by ultrashort pulse X-rays.

[0008] The technical solution of this application is implemented as follows: In a first aspect, this application provides a magnetic particle acoustic imaging system excited by ultrashort pulse X-rays, comprising: An X-ray signal emission module is used to respond to control commands and control an X-ray generator to generate X-ray pulses with preset parameters, so that the X-ray pulses act on a pre-set magnetic particle marking region, wherein the magnetic particles include a magnetic core, a functional modification layer and a biocompatible matrix, and the magnetic particle ligands include targeting molecules. A magnetic field modulation module is used to modulate the magnetic particle marking area so that when the X-ray pulse acts on the magnetic particles, it generates an original acoustic signal that matches the magnetic field modulation. An acoustic detection module is used to convert the raw acoustic signal into a raw data signal; The signal processing module is used to receive the raw data signal and process the raw data signal to obtain the imaging result.

[0009] In one specific embodiment, the X-ray signal transmitting module includes: A position control unit is used to receive external feedback information and adjust the X-ray generator to the corresponding position according to the external feedback information. The signal transmitting unit is used to control the X-ray generator to generate an X-ray pulse with preset parameters when the corresponding position coincides with the preset magnetic particle marking area, wherein the preset parameters include tube voltage, pulse width, and pulse frequency.

[0010] In one specific embodiment, the magnetic field modulation module includes: A static magnetic field modulation unit is used to generate a magnetic field signal of fixed intensity to magnetize magnetic particles, so that the magnetic moments of the magnetic particles are aligned in a direction that matches the magnetic field signal of fixed intensity. A dynamic magnetic field modulation unit is used to generate a gradient magnetic field signal based on the fixed-intensity magnetic field signal so that the fixed-intensity magnetic field signal and the gradient magnetic field signal form a composite magnetic field signal for encoding the composite signal.

[0011] In one specific embodiment, the dynamic magnetic field modulation unit includes: A waveform generator is used to generate corresponding alternating signals according to frequency commands. A gradient amplifier is used to output an adjustable drive current signal according to the alternating signal, so as to generate a corresponding gradient magnetic field signal through a gradient coil.

[0012] In one specific embodiment, a signal synchronization module is further included, which is used to send a synchronization trigger signal to the magnetic field modulation module in response to the X-ray signal transmitting module sending an X-ray pulse, so that the signal frequencies of the X-ray signal transmitting module and the magnetic field modulation module are matched.

[0013] In one specific embodiment, the signal processing module includes: The acoustic signal extraction unit is used to acquire raw data signals and process them to obtain relaxation correction signals. An image reconstruction unit is used to convert the relaxation correction signal into a magnetic particle distribution image and a concentration distribution image; The cell state recognition unit is used to determine the attenuation threshold classification based on the acoustic impedance of different tissues to determine the cell state based on the attenuation threshold classification, to label the cell state to the magnetic particle distribution image to generate a cell map with state labels, and to determine quantitative concentration data based on the concentration distribution image and the cell state. An image fusion unit is used to fuse a cell image with status labels, concentration quantification data, and an anatomical structure image corresponding to the cell image with status labels to obtain a fused image, wherein the anatomical structure image includes a tissue structure image and a microvascular imaging image.

[0014] In one specific embodiment, the acoustic signal extraction unit includes: A noise suppression subunit is used to filter the original acoustic image according to a predetermined filtering threshold to obtain a denoised signal, wherein the filtering threshold is determined based on the noise power within the signal frequency band; A signal enhancement subunit is used to perform signal gain on the denoised signal to amplify the signal intensity in the high gradient region to obtain an enhanced denoised signal. The signal averaging and sensitivity enhancement subunit is used to coherently superimpose and average the amplified enhanced and denoised signal, and then filter the effective data to obtain the average sensitivity enhancement signal. The relaxation correction subunit is used to deconvolve the average sensitivity enhancement signal in the frequency domain to obtain the relaxation correction signal.

[0015] Secondly, this application provides a method for magnetic particle acoustic imaging excited by ultrashort pulse X-rays, comprising: In response to a control command, the X-ray generator is controlled to generate an X-ray pulse with preset parameters, such that the X-ray pulse acts on a pre-set magnetic particle-marked region, wherein the magnetic particles include a magnetic core, a functional modification layer and a biocompatible matrix, and the magnetic particle ligands include a targeting molecule. The magnetic particle-marked region is modulated with a magnetic field so that when the X-ray pulse acts on the magnetic particles, it generates an original acoustic signal that matches the magnetic field modulation. The original acoustic signal is converted into a raw data signal, and the raw data signal is processed to obtain the imaging result.

[0016] In one specific embodiment, processing the raw data signal to obtain the imaging result includes: The raw data signal is acquired and processed to obtain the relaxation correction signal; The relaxation correction signal is converted into a magnetic particle distribution image and a concentration distribution image; The attenuation threshold is determined based on the acoustic impedance of different tissues to classify cell states, and the cell states are labeled onto the magnetic particle distribution image to generate a cell map with state labels. Concentration quantitative data is determined based on the concentration distribution image and the cell states. A fused image is obtained by fusing a state-labeled cell map, quantitative concentration data, and an anatomical structure map corresponding to the state-labeled cell map, wherein the anatomical structure map includes a tissue structure map and a microvascular imaging map.

[0017] In one specific implementation, the raw data signal is acquired and processed to obtain a relaxation correction signal, including: The original acoustic image is filtered according to a predetermined filtering threshold to obtain a denoised signal, wherein the filtering threshold is determined based on the noise power within the signal frequency band; The denoised signal is amplified by signal gain to increase the signal intensity in the high gradient region, thereby obtaining an enhanced denoised signal. The amplified and denoised signal is coherently superimposed and averaged, and then the effective data is filtered to obtain the average sensitivity-enhancing signal. The relaxation correction signal is obtained by deconvolving the average sensitivity enhancement signal in the frequency domain.

[0018] Thirdly, embodiments of this application provide a magnetic particle acoustic imaging device excited by ultrashort pulse X-rays, the device comprising: a processor and a memory; wherein, The memory is used to store computer programs; The processor is configured to call and run the computer program from the memory to perform the method as described in the second aspect.

[0019] Fourthly, embodiments of this application provide a computer-readable storage medium storing executable instructions for causing a processor to perform the method described in the second aspect.

[0020] This application provides a method and system for magnetic particle acoustic imaging excited by ultrashort pulse X-rays, with a penetration depth ≥15cm (covering all deep tissues of the human body), spatial resolution ≤250μm (based on formula calculation and experimental verification), temporal resolution ≤100μs / frame, sensitivity reaching the level of hundreds of cells (ng magnetic particles), single scan radiation dose ≤0.1mSv, and annual radiation dose below the safety threshold of 5mSv, meeting the needs of long-term monitoring. The precise magnetic field modulation module adopts a low-power architecture, reducing the total energy consumption of the device by more than 40%. Through "static + dynamic gradient" composite magnetic field modulation, the ratio of magnetic particle signal to background signal is >50:1, effectively distinguishing between live cells and free magnetic particles and avoiding misjudgment; the image fusion error is ≤20μm, and the cell position is accurately matched with the surrounding anatomical structure, providing complete information for diagnosis and treatment; the device is small in size, low in cost, and low in energy consumption, suitable for promotion in primary hospitals, and can be expanded to multiple scenarios such as stem cell tracing, CAR-T monitoring, microvascular imaging, and surgical navigation, with broad clinical application value. Attached Figure Description

[0021] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the specification, serve to explain the technical solutions of this application. Obviously, the drawings described below are merely some embodiments of this application, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.

[0022] The flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all content and operations / steps, nor do they necessarily have to be performed in the described order. For example, some operations / steps can be broken down, while others can be combined or partially combined; therefore, the actual execution order may change depending on the specific circumstances.

[0023] Figure 1 An optional block diagram of an ultrashort pulse X-ray excited magnetic particle acoustic imaging system provided in an embodiment of this application; Figure 2 A schematic flowchart of a magnetic particle acoustic imaging method excited by ultrashort pulse X-rays is provided for an embodiment of this application. Figure 3 This is a schematic diagram of the structure of a magnetic particle acoustic imaging device excited by ultrashort pulse X-rays, provided in an embodiment of this application. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the specific technical solutions of this application will be further described in detail below with reference to the accompanying drawings of the embodiments of this application. The following embodiments are used to illustrate this application, but are not intended to limit the scope of this application.

[0025] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0026] In the following description, references to "some embodiments," "this embodiment," "this application embodiment," and examples, etc., describe a subset of all possible embodiments. However, it is understood that "some embodiments" may be the same subset or different subset of all possible embodiments and may be combined with each other without conflict.

[0027] If the application documents contain similar descriptions such as "first / second", the following explanation shall be added: In the following description, the terms "first / second / third" are used only to distinguish similar objects and do not represent a specific order of objects. It is understood that "first / second / third" may be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0028] This application provides a magnetic particle acoustic imaging system excited by ultrashort pulse X-rays. Figure 1 An optional module block diagram of an ultrashort pulse X-ray excited magnetic particle acoustic imaging system provided in this application embodiment will be combined with Figure 1 The steps shown are explained.

[0029] A magnetic particle acoustic imaging system excited by ultrashort pulse X-rays, comprising: X-ray signal emission module 1 is used to respond to control commands to control the X-ray generator to generate X-ray pulses with preset parameters, so that the X-ray pulses act on a pre-set magnetic particle marking area, wherein the magnetic particles include a magnetic core, a functional modification layer and a biocompatible matrix, and the magnetic particle ligands include targeting molecules. This embodiment first describes the main hardware and principles of the X-ray signal transmission module. The X-ray generator produces ultrashort pulse X-rays through an X-ray tube. In this embodiment, an ultrashort pulse refers to X-rays with a pulse width ≤ 100 ns, which can be generated by an X-ray tube of 40k~200kVp. In this embodiment, after the X-rays penetrate human tissue, they are absorbed by magnetic particles (SPIONs, with a particle size of, for example, 10-20 nm). According to Beer-Lambert's law, the X-ray intensity attenuates as I = I0exp(-μ×x) (where μ is the linear attenuation coefficient and x is the penetration depth), and the energy E absorbed by the magnetic particles is... abs =I0(1-exp(-μ SPION ×d))A (where μ SPION The mass attenuation coefficient of the magnetic particles is experimentally calibrated to be 5.8 cm. - 2kg - ¹@100keV, where d is the characteristic size of the magnetic particle and A is the absorption cross-section. The absorbed energy is converted into heat energy in an extremely short time (e.g., nanoseconds), causing a transient temperature rise ΔT in the magnetic particle and its adjacent medium, with a rate that can reach 10⁻⁶. 5 Temperatures above °C / s, but due to the extremely short pulse and small total heat, the peak temperature rise is controllable (<1°C), thus avoiding thermal damage. For scenarios requiring precise control, calculations can be performed using formulas. The temperature gradient formula is ΔT = Q / c × ρ × r. 3 Where Q is the absorbed X-ray energy, c is the specific heat capacity of the medium, ρ is the density of the medium, and r is the radius of the magnetic particle.

[0030] In addition, radiation dose data can be collected in real time and fed back to the control host. Based on the radiation dose formula D=k×V×I×t (where k is the dose coefficient, V is the tube voltage, I is the tube current, and t is the scanning time), the radiation dose of a single scan is ensured to be ≤0.1mSv.

[0031] Control commands can be, for example, commands sent by the host computer, which set parameters such as tube voltage, pulse width, pulse frequency, and beam diameter.

[0032] It should be noted that the magnetic particle labeling region needs to be determined in advance, depending on the specific protocol. First, the magnetic particles need to be prepared. Based on the preparation parameters, magnetic particles (Fe3O4 nanoparticles, CD44 / CD3 antibodies, and other targeting molecules) with appropriate particle size and modification ratios are selected, ultimately yielding magnetic particles (SPIONs) with surface-modified targeting peptides. The characteristics of the magnetic particles can be determined using parameters such as particle size and magnetic saturation intensity. Second, the magnetically labeled cells are delivered to the target location using an integrated microinjector and optical positioning system, and the coordinates and other parameters of the target location are recorded. The magnetic particle labeling region can then be determined based on these coordinate parameters.

[0033] In one specific embodiment, the X-ray signal transmitting module includes: A position control unit is used to receive external feedback information and adjust the X-ray generator to the corresponding position according to the external feedback information. Specifically, a 3D map of the position control unit can be constructed, and the coordinate information fed back from the outside can be mapped onto the 3D map to control the X-ray transmitter to move to the corresponding position. The external feedback information can be pre-set and stored data parameters, or it can be obtained in real time by converting external input parameter information.

[0034] The signal transmitting unit is used to control the X-ray generator to generate an X-ray pulse with preset parameters when the corresponding position coincides with the preset magnetic particle marking area, wherein the preset parameters include tube voltage, pulse width, and pulse frequency.

[0035] The signal transmitting unit in this embodiment can be set up independently or integrated and fixed with devices such as ultrasonic probes or magnetic control coils. When integrating and fixing, different positional relationships can be set according to different usage requirements. Of course, regardless of the fixing method, a three-dimensional map needs to be constructed based on the relative positions in order to accurately adjust the position.

[0036] In summary, this embodiment optimizes the energy (controlled by the X-ray tube voltage of 40k~200kVp) and pulse frequency of ultrashort pulse X-rays. Based on the radiation dose formula D=k×V×I×t (where k is the dose coefficient, V is the tube voltage, I is the tube current, and t is the scan time), it controls the radiation dose of a single scan to ≤0.1mSv (only 1 / 50 of that of CT). It adopts a low-power precision magnetic field modulation architecture to replace the traditional high-power magnetic field system, which can reduce the size of the equipment and save equipment costs, meeting the needs of clinical application.

[0037] A magnetic field modulation module is used to modulate the magnetic particle marking area so that when the X-ray pulse acts on the magnetic particles, it generates an original acoustic signal that matches the magnetic field modulation. This embodiment explains the principles of magnetic field modulation and acoustic signal generation as follows: Magnetic field modulation is primarily based on the formula for magnetic moment orientation alignment: M = M0 × (B / (k × T)) (where M is the magnetization, M0 is the saturation magnetization, B is the static magnetic field strength, k is the Boltzmann constant, and T is the absolute temperature). For a static bias magnetic field, the strength is typically 0-10 mT, used to magnetize SPIONs (magnetic particles), orienting their magnetic moments to change their X-ray absorption cross-section (magnetostrictive absorption effect), thereby improving energy absorption efficiency under specific polarization directions. For a dynamic gradient magnetic field, the strength is typically 0-1 mT / m, and the frequency is typically 1-100 kHz. The dynamic gradient magnetic field enables precise spatial modulation of the magnetic field strength. Its gradient direction is synchronized with the X-ray excitation region scan, further amplifying the thermoacoustic response differences of the magnetic particles in the target region and suppressing background noise through spatial coding.

[0038] When an X-ray pulse acts on magnetic particles, it generates an acoustic signal through the thermoelastic effect. The instantaneous temperature rise ΔT of the magnetic particles in a localized area induces thermal expansion and contraction of the surrounding medium (such as intracellular fluid and interstitial fluid), based on the thermoelastic equation. ²u–( ) ²u / t² = (β / C p ) T / t (where u is the displacement vector, β is the volume expansion coefficient, C) p Given the isobaric specific heat capacity, where T is temperature and t is time, a pressure wave p(r,t) is generated, which is the acoustic signal. Under stress constraint conditions, the initial pressure p0 is proportional to the absorbed energy density: p0 ≈ Γ × η th ×μ a ×F, where Γ is the Grindelwald parameter, η th It is thermal efficiency, μ a α is the absorption coefficient, and F is the luminous flux. Magnetic particles within living cells are constrained by the cell membrane and organelles, resulting in different thermal diffusion and acoustic coupling characteristics compared to free magnetic particles. This leads to characteristic differences in the center frequency, bandwidth, and attenuation rate of the acoustic signals they generate. By analyzing the attenuation coefficient α (unit: dB / mm), the acoustic signal attenuation rate is significantly lower than that of free magnetic particles, thus distinguishing the two.

[0039] In other words, a continuously applied static bias magnetic field can increase the detectable range of acoustic signals, while a dynamic gradient magnetic field can accurately detect and track the specific distribution characteristics of magnetic particles within the detection range.

[0040] Specifically, the magnetic field modulation module includes: A static magnetic field modulation unit is used to generate a magnetic field signal of fixed intensity to magnetize magnetic particles, so that the magnetic moments of the magnetic particles are aligned in a direction that matches the magnetic field signal of fixed intensity. A dynamic magnetic field modulation unit is used to generate a gradient magnetic field signal based on the fixed-intensity magnetic field signal so that the fixed-intensity magnetic field signal and the gradient magnetic field signal form a composite magnetic field signal, thereby suppressing background noise through spatial coding.

[0041] For example, a neodymium iron boron permanent magnet (nickel-copper-nickel coating, size 200×50×15mm³) can be used for the static magnetic field modulation unit. The host control unit can output a uniform static magnetic field that is adjustable from 0 to 10mT (uniformity in the imaging area <5%) by inputting a magnetic field strength calibration signal, so that the magnetic moments of the magnetic particles are oriented.

[0042] The dynamic gradient magnetic field unit can, for example, use multiple pairs of Golay-type transverse gradient coils (x / y / z directions), with each pair of coils having 200 turns. The host control makes the output a 0-1mT / m linear gradient magnetic field by controlling the drive current (20-60A) of the gradient amplifier. The gradient direction and magnitude are independently controlled by the current of each pair of coils, realizing spatially precise encoding of the magnetic field strength.

[0043] To effectively control the dynamic gradient magnetic field unit, the dynamic magnetic field modulation unit may include: A waveform generator is used to generate corresponding alternating signals according to frequency commands. A gradient amplifier is used to output an adjustable drive current signal according to the alternating signal, so as to generate a corresponding gradient magnetic field signal through a gradient coil.

[0044] Specifically, the gradient amplifier (with the alternating signal from the waveform generator as input and the adjustable drive current of 0-100A as output) and the waveform generator (with the frequency command from the control host as input and the cosine alternating signal of 1-5kHz as output) can adjust the magnetic field modulation parameters (intensity, frequency, and gradient direction) in real time.

[0045] To ensure the coordination between magnetic field modulation and X-ray excitation, a signal synchronization module may also be included, which, in response to the X-ray signal transmitting module sending an X-ray pulse, sends a synchronization trigger signal to the magnetic field modulation module so that the signal frequencies of the X-ray signal transmitting module and the magnetic field modulation module are matched.

[0046] This embodiment improves the temporal resolution to ≤100μs / frame by rapidly exciting ultrashort pulse X-rays, amplifying the signal through precise magnetic field modulation, and real-time detection of acoustic signals. Based on the linear relationship between magnetic particle signal intensity and quantity, S=k×N×M (where S is the acoustic signal amplitude, k is the system response coefficient, N is the number of cells, and M is the magnetic particle load in a single cell), combined with a signal superposition algorithm, the sensitivity is improved to detect ≤100 magnetically labeled cells (corresponding to ng-level magnetic particle mass, specifically determined experimentally based on SPION particle size, magnetization intensity, and labeling efficiency), enabling real-time monitoring of cell dynamic migration.

[0047] An acoustic detection module is used to convert the raw acoustic signal into a raw data signal; The signal processing module is used to receive the raw data signal and process the raw data signal to obtain the imaging result. In this embodiment, the original acoustic signal can be received by an array of ultrasonic transducers.

[0048] Specifically, the array ultrasound transducer can be a multi-element linear array transducer, with the input being the raw acoustic signal generated by magnetic particles and the tissue ultrasound signal, and the output being an analog electrical signal; the center frequency is 20MHz (adjustable range 0-50MHz), and the receiving sensitivity is ≤10. - ¹³Pa, detection depth ≥15cm, and the angle can be adjusted (0-90°) by a robotic arm to achieve multi-angle signal acquisition.

[0049] Preferably, it also includes an anti-interference circuit to reduce electromagnetic interference between the X-ray generator and the dynamic gradient magnetic field unit and the static magnetic field modulation unit, so as to improve the effectiveness of the acquired data.

[0050] This embodiment utilizes the deep penetration capability of ultrashort pulse X-rays (pulse width ≤ 1 ns) and the signal enhancement effect of precise magnetic field modulation, combined with the high resolution advantage of acoustic imaging, to achieve an imaging penetration depth ≥ 15 cm and a spatial resolution ≤ 250 μm (based on the spatial resolution formula Δs = 0.8 × v). s / f max , where v s The average speed of sound in the organization is approximately 1540 m / s, f max The highest receiving frequency of the ultrasonic transducer, for example, f. max At 5MHz, the theoretical limit resolution is Δs≈250μm, which meets the cell tracing requirements of deep tissues such as the brain and spinal cord.

[0051] Specifically, the signal processing module includes: The acoustic signal extraction unit is used to acquire raw data signals and process them to obtain relaxation correction signals. For the acquisition of analog electrical signals, a multi-channel (e.g., 16-channel) high-speed data acquisition card is used. The analog electrical signals and synchronous trigger signals are received and acquired synchronously, then processed to output digital signals.

[0052] The processing of acoustic signals can be carried out through the following steps: A noise suppression subunit is used to filter the original acoustic image according to a predetermined filtering threshold to obtain a denoised signal, wherein the filtering threshold is determined based on the noise power within the signal frequency band; A wavelet threshold filtering algorithm (using the db6 wavelet basis) is employed, with a digitized signal as input and a denoised magnetic particle acoustic signal as output; based on the signal-to-noise ratio formula SNR=10lg (P signal / P noise (where P) signal For signal power, P noise (Noise power), the signal-to-noise ratio was improved to ≥30dB, tissue self-acoustic signals (such as low-frequency signals generated by vascular pulsation) and X-ray scattering noise were removed, and magnetic particle acoustic signals of 0-50MHz were retained.

[0053] A signal enhancement subunit is used to perform signal gain on the denoised signal to amplify the signal intensity in the high gradient region to obtain an enhanced denoised signal. The input is the spatial coding information of the denoised signal and magnetic field modulation, and the output is the enhanced signal. By using the spatial coding weight of the dynamic gradient magnetic field, the signal in different regions is amplified in a targeted manner, so that the ratio of magnetic particle signal to background signal is increased to >50:1, thereby enhancing the signal specificity.

[0054] Preferably, cell viability can also be distinguished based on this. For example, the input is the enhanced signal, and the output is the cell viability determination result. Based on the attenuation coefficient formula α=-(1 / d) lg (A / A0) (where d is the propagation distance, A is the signal amplitude after propagation, and A0 is the initial signal amplitude), the attenuation coefficient α (dB / mm) of the acoustic signal is extracted. The attenuation coefficient is determined by a threshold (α≤0.8dB / mm indicates viable cells and magnetic particles, α>1.0dB / mm indicates ionized magnetic particles), and different colors are used to mark the areas in the image (green for viable cell areas and yellow for ionized magnetic particles).

[0055] The signal averaging and sensitivity enhancement subunit is used to coherently superimpose and average the amplified enhanced and denoised signals, then filter the effective data to obtain the average sensitivity enhancement signal. The input is the enhanced signal from multiple excitations, and the output is the superimposed signal. A signal superposition algorithm is used to superimpose and average the acoustic signals generated by the combined effect of 10 consecutive X-ray excitations and magnetic field modulation in the same region, based on the signal superposition formula S. sum =∑S i (i=1 to 10, S) i (for single signal amplitude), so that the detection limit reaches the level of hundreds of cells (ng magnetic particles).

[0056] The relaxation correction subunit is used to deconvolve the average sensitivity enhancement signal in the frequency domain to obtain the relaxation correction signal; the input is the superimposed signal, and the output is the relaxation correction signal; a "relaxation effect deconvolution" module is introduced, based on the deconvolution formula S corr =S measured h - ¹ (where S) corr To correct the signal, S measured (where h is the measurement signal and h is the relaxation response function), correcting the signal hysteresis caused by the relaxation of magnetic moments of magnetic particles, and further improving the temporal resolution of dynamic imaging.

[0057] An image reconstruction unit is used to convert the relaxation correction signal into a magnetic particle distribution image and a concentration distribution image; The image reconstruction process can be carried out through the following steps: a. Cell Distribution Image Reconstruction: The input consists of the corrected signal, the position of the ultrasonic transducer elements, and the magnetic field coding information; the output is a cell distribution image. Combining the position of the ultrasonic transducer elements, the signal arrival time, and the spatial coding information of the magnetic field modulation, the image is reconstructed based on the back-projection algorithm and the three-dimensional coordinate calculation formula (x,y,z)=f(t1,t2,...,t...). n (where t) i The time it takes for the signal to reach each array element is used to generate a cell distribution image with a spatial resolution of ≤250μm; the temporal resolution is controlled by the coordinated control of the X-ray pulse frequency and the magnetic field modulation frequency (10kHz pulse frequency corresponds to 100μs / frame), which can track cell migration in real time (such as the movement of T cells to the tumor lesion).

[0058] b. Tissue Structure Image Reconstruction: The input is an ultrasound signal, and the output is a tissue structure image; the signal received by the ultrasound transducer is subjected to beamforming (using a delay-addition algorithm) and grayscale conversion, based on the grayscale value formula G=255×(A / A). max (where A is the signal amplitude, A) max(Maximum amplitude), generating tissue structure images (such as sulci and gyri of the brain, gray matter columns of the spinal cord), with a spatial resolution ≤250μm.

[0059] An image fusion unit is used to fuse a cell image with status labels, concentration quantification data, and an anatomical structure image corresponding to the cell image with status labels to obtain a fused image, wherein the anatomical structure image includes a tissue structure image and a microvascular imaging image.

[0060] Image fusion: The input consists of a cell distribution image and a tissue structure image, and the output is a fused image. A pixel-level fusion algorithm is used, based on the fusion formula F(x,y)=w1×C(x,y)+w2×T(x,y) (where F(x,y) is the fused pixel value, C(x,y) is the pixel value of the cell distribution image, T(x,y) is the pixel value of the tissue structure image, and w1 and w2 are weighting coefficients, with a sum of 1). The cell distribution image and the tissue structure image are superimposed to make the cell position accurately match the surrounding anatomical structure (such as tumor boundary and blood vessels). The fusion error is ≤20μm, which makes it easier for doctors to judge the spatial relationship between cells and lesions.

[0061] Please continue reading Figure 2 This embodiment also provides a method for magnetic particle acoustic imaging excited by ultrashort pulse X-rays, including: In response to a control command, the X-ray generator is controlled to generate an X-ray pulse with preset parameters, such that the X-ray pulse acts on a pre-set magnetic particle-marked region, wherein the magnetic particles include a magnetic core, a functional modification layer and a biocompatible matrix, and the magnetic particle ligands include a targeting molecule. The magnetic particle-marked region is modulated with a magnetic field so that when the X-ray pulse acts on the magnetic particles, it generates an original acoustic signal that matches the magnetic field modulation. The original acoustic signal is converted into a raw data signal, and the raw data signal is processed to obtain the imaging result.

[0062] Specifically, the original data signal is processed to obtain the imaging result, including: The raw data signal is acquired and processed to obtain the relaxation correction signal; The relaxation correction signal is converted into a magnetic particle distribution image and a concentration distribution image; The attenuation threshold is determined based on the acoustic impedance of different tissues to classify cell states, and the cell states are labeled onto the magnetic particle distribution image to generate a cell map with state labels. Concentration quantitative data is determined based on the concentration distribution image and the cell states. A fused image is obtained by fusing a state-labeled cell map, quantitative concentration data, and an anatomical structure map corresponding to the state-labeled cell map, wherein the anatomical structure map includes a tissue structure map and a microvascular imaging map.

[0063] Specifically, the raw data signal is acquired and processed to obtain the relaxation correction signal, including: The original acoustic image is filtered according to a predetermined filtering threshold to obtain a denoised signal, wherein the filtering threshold is determined based on the noise power within the signal frequency band; The denoised signal is amplified by signal gain to increase the signal intensity in the high gradient region, thereby obtaining an enhanced denoised signal. The amplified and denoised signal is coherently superimposed and averaged, and then the effective data is filtered to obtain the average sensitivity-enhancing signal. The relaxation correction signal is obtained by deconvolving the average sensitivity enhancement signal in the frequency domain.

[0064] The following is a further example: I. Preparation and Characterization of Magnetic Particles The magnetic particles used in this embodiment are ABS nanocomposites with a "magnetic core-functional modification layer-biocompatible matrix" structure. The specific preparation and characterization are as follows: 1. Synthesis of ABS Nanocomposites ① Synthesis of albumin-bismuth (AB) complex: 13.2g BSA was dissolved in 200mL of double-distilled water, and nitric acid solution containing 17mg bismuth nitrate pentahydrate was added dropwise. NaOH solution was added under stirring to adjust the pH to 12.75. After stirring for 8 hours, the mixture was centrifuged and washed to obtain the AB complex. ② Synthesis of ABS complex: 6 mg EDAC was mixed with 2 mL SPIO (1 mg Fe / mL), 12 mL double-distilled water was added and shaken for 15 min, then 2 mL AB complex (5 mg Bi / mL) was added and shaken for 2 h. After purification by centrifugation at 6400×g for 5 min, the mixture was ultrasonically dispersed for 15 min to obtain ABS nanocomposite.

[0065] 2. Key characterization data ① Element ratio: Bi:Fe = 5:1 was determined by ICP-MS (this is the optimal ratio for X-ray attenuation and magnetic response characteristics, verified by phantom experiments). ② Particle size and magnetic properties: The magnetic core SPIONs have a particle size of 10-20 nm and a saturation magnetization of ≥75 emu / g (measured by a vibrating sample magnetometer VSM). The overall particle size of the ABS composite is 86-102 nm and the surface ζ potential is -23 to -24 mV (measured by a laser particle size analyzer and potentiometer). ③Biocompatibility: According to ISO 10993 biocompatibility test, the LD50 in mouse model is >500mg / kg. Long-term metabolic test shows that more than 90% of the magnetic particles are excreted through liver metabolism and there is no accumulation in the body (verified by pathological sections and inductively coupled plasma mass spectrometry). ④ Labeling efficiency: After covalently binding specific antibodies (such as stem cell CD133, T cell CD3) to the surface, the labeling efficiency of the target cells was verified by flow cytometry to be ≥95%, and the cell viability was verified by MTT assay to be >95%.

[0066] II. Detailed Implementation Instructions for Each Module The ultrashort pulse X-ray excited magnetic particle acoustic imaging system of this application includes an X-ray signal transmission module, a magnetic field modulation module, an acoustic detection module, and a signal processing module. For complete illustration, it may also include a magnetic particle label delivery unit, a target area focusing and positioning module, and a system control and display module. The modules work together and are synchronized at the nanosecond level by FPGA. The signal flow of the overall system is as follows: magnetic particle labeling → target area positioning → X-ray pulse excitation + magnetic field composite modulation → magnetic particle thermoacoustic effect sound generation → acoustic signal acquisition and processing → image reconstruction and fusion → result display and storage.

[0067] For the target area focusing and positioning module, this module is the foundation of imaging accuracy. It adopts the design of "motor-driven mechanical structure + laser positioning calibration + closed-loop feedback". Its core function is to achieve precise coaxial alignment of the X-ray focal point, magnetic field modulation area and ultrasound focal point, with a positioning deviation of ≤80μm (50μm magnetic particle point source implanted in the phantom, repeated positioning 10 times, average deviation 62μm, maximum deviation 78μm).

[0068] 1. Components: X / Y / Z three-axis stepper motor (model: 42HS48-1684, stepping accuracy ≤10μm, X-axis travel 0-80cm, Y-axis 0-50cm, Z-axis 50-100cm), 650nm semiconductor laser emitter (coaxiality error with X-ray focusing beam ≤50μm), high-definition vision sensor (1920×1080 resolution, frame rate 30fps), Hall effect magnetic field sensor array (resolution 0.1mT, sampling rate 1kHz), ultrasonic transducer micro positioning chip (accuracy ≤5μm); 2. Implementation process: ①Import the three-dimensional coordinates (x0, y0, z0) of the target area for magnetic particle delivery, and control the host to drive the three-way motor to adjust the position of the X-ray generator; ② The laser emitter emits a positioning laser, and the visual sensor collects the deviation between the laser spot and the positioning mark on the target location. Combined with the coordinate data from the magnetic field sensor and the ultrasonic positioning chip, the system is dynamically calibrated using a PID algorithm. ③ After calibration, a "positioning ready" signal is sent back to the control host to trigger the operation of subsequent modules. The closed-loop feedback throughout the process ensures positioning accuracy.

[0069] As for the X-ray signal emission module, this module is used to generate ultrashort pulse X-rays with preset parameters, which precisely excites the magnetic particles in the magnetic particle-marked area to generate heat, ensuring deep penetration while keeping the radiation dose controllable.

[0070] 1. Components: Position control unit, signal transmission unit (40k~200kVp adjustable miniature pulse X-ray tube, collimator, dose monitoring unit), and synchronization trigger subunit; 2. Core parameters and implementation basis: ① X-ray pulse parameters: pulse width ≤ 100ns, tube voltage 40k~200kVp (dynamically adjusted according to the imaging scenario, such as 80kVp for brain stem cell tracing and 150kVp for lung cancer CAR-T monitoring), pulse frequency 1-100kHz, beam diameter 1-5mm (collimator control); ② Radiation dose control: Based on the radiation dose formula D=k×V×I×t (k is the dose coefficient, calibrated to 0.0025mSv / (kV)). mA The error is ≤ ±5%; V is the tube voltage, kVp; I is the tube current, mA; t is the scan time, s). The tube current is fixed at 1 mA, the scan time is controlled at 7-9 s, the radiation dose per scan is ≤ 0.1 mSv (data is collected and fed back in real time by the dose monitoring unit, the actual measured dose for brain tracing is 0.08 mSv, and the actual measured dose for lung cancer tracing is 0.09 mSv), and the annual radiation dose is below the safety threshold of 5 mSv. ③ Position and emission control: The position control unit receives coordinate feedback from the focusing and positioning module. After confirming that the X-ray generator coincides with the target area (deviation ≤ 80μm), the signal emission unit controls the X-ray tube to generate X-ray pulses, and the synchronous triggering subunit sends synchronous trigger signals to the magnetic field modulation module and the acoustic detection module.

[0071] 3. The principle of heat generation by magnetic particles and experimental evidence: X-rays are absorbed by magnetic particles after penetrating human tissue, following the principle of magnetic absorption. Lambert's Law: I = I0exp(-μ×x) (μ is the linear decay coefficient of the tissue; for soft tissue, μ≈0.15cm) - ¹@100keV; x is the penetration depth), magnetic particle absorbs energy E abs =I0(1-exp(-μ SPION ×d))A(μ SPION The mass attenuation coefficient of the magnetic particles is experimentally calibrated to 5.8 cm. - 2kg - ¹@100keV; d is the characteristic size of the magnetic particle (15nm); A is the absorption cross-section (2.3×10⁻⁶). - ¹ 4 cm²); The absorbed energy is converted into heat energy in nanoseconds, and the transient temperature rise follows the temperature gradient formula ΔT=Q / c×ρ×r³ (where Q is the absorbed X-ray energy; c is the specific heat capacity of the medium, 4.18 J / (g)). (℃); ρ is the medium density 1.0 g / cm³; r is the magnetic particle radius), the experimentally determined transient temperature rise is <1°C, avoiding thermal damage to biological tissues.

[0072] For the magnetic field modulation module, this module achieves magnetic particle signal enhancement and spatial coding through "static + dynamic gradient" composite magnetic field modulation. The core principle is based on the magnetic moment orientation arrangement formula. The parameters have been experimentally verified to ensure that the signal-to-background ratio is >50:1, laying the foundation for the specific extraction of acoustic signals.

[0073] 1. Components: Static magnetic field modulation unit, dynamic magnetic field modulation unit (waveform generator, gradient amplifier, Golay type triaxial gradient coil, relaxation correction unit), and signal synchronization module; 2. Implementation details for each unit: ① Static magnetic field modulation unit: NdFeB permanent magnets (200×50×15mm³) with nickel-copper-nickel coating are used to generate a uniform static magnetic field that is adjustable from 0 to 10mT. The magnetic field uniformity is <5% (the position of the permanent magnet is monitored and dynamically adjusted in real time). Based on the magnetic moment orientation formula M=M0×(B / (k×T)) (M is the magnetization intensity, M0 is the saturation magnetization intensity, B is the static magnetic field intensity, k is the Boltzmann constant, and T is the absolute temperature), the magnetic moments of the magnetic particles are oriented, and the acoustic signal intensity is improved by more than 30%.

[0074] ② Dynamic magnetic field modulation unit: It adopts three pairs of Golay type transverse gradient coils (x / y / z directions, 200 turns per pair). The waveform generator generates a 1-5kHz cosine alternating signal according to the frequency command. The gradient amplifier converts it into an adjustable drive current of 20-60A, generating a linear dynamic gradient magnetic field of 0~1mT / m. The gradient direction is synchronized with the X-ray excitation area scanning. The signal in the target area is amplified in a targeted manner by the spatial coding formula ω=k×G (k is the system weight coefficient, G is the gradient intensity) to suppress background noise. ③ Relaxation correction unit: based on the deconvolution formula S corr =S measured h - ¹(S corr To correct the signal, S measured (where h is the relaxation response function), corrects the signal lag caused by magnetic particle moment relaxation, with a timing error ≤10ns, improving the temporal resolution of dynamic imaging; ④ Signal Synchronization Module: Responds to the synchronous trigger signal of X-ray emission and sends trigger signals to the magnetic field modulation module and acoustic detection module. The synchronization error is ≤10ns. The triggering sequence is as follows: 0ns after X-ray pulse emission, the magnetic field gradient switching is triggered, and 5ns after the ultrasound acquisition window is opened to ensure that the three work together.

[0075] (iv) Magnetic particle tag delivery unit This module is used to precisely deliver ABS nanocomposites to target cells and complete labeling, providing a tracer basis for imaging, with a delivery error ≤30μm and labeling efficiency ≥95%.

[0076] 1. Components: Micro-injection module (10μL micro-injector, adapted for cell / magnetic particle delivery), optical positioning sub-module (shares a vision sensor with the focusing positioning module to calibrate the delivery position). 2. Implementation process: ① Select delivery method according to imaging scenario: Stem cells / tumor local cells are delivered by stereotactic micro-injection, peripheral blood immune cells are delivered by intravenous infusion, and wound blood vessels are delivered by subcutaneous multi-point injection; ② The optical positioning submodule calibrates the delivery position in real time to ensure that the delivery error is ≤30μm; ③ Co-culture of magnetic particles with target cells (stem cells for 24 hours, T cells for 16 hours), flow cytometry confirmed labeling efficiency ≥95%, and transmission electron microscopy showed that the magnetic particles were uniformly distributed in the cells without obvious aggregation.

[0077] (v) Acoustic signal detection and signal processing module Acoustic signal detection is performed by an array of ultrasonic transducers. The signal processing module performs "denoising, enhancement, correction, and feature analysis" on the acquired acoustic signals, ultimately achieving cell state differentiation and image reconstruction, ensuring signal specificity and imaging accuracy.

[0078] 1. Acoustic signal detection unit Employing a multi-element linear array ultrasonic transducer, with core parameters: center frequency 20MHz (adjustable from 0-50MHz), receiving sensitivity ≤10. - ¹³Pa, detection depth ≥15cm, angle adjustable from 0-90° via robotic arm; equipped with a multi-channel high-speed data acquisition card (sampling rate 100MS / s, resolution 16bit), built-in anti-interference circuit (shielding ground, filter capacitor, differential transmission) to reduce electromagnetic interference from X-rays and magnetic fields, and convert analog acoustic signals into digital signals.

[0079] 2. Core Signal Processing Flow This embodiment employs a four-step signal processing strategy: two-stage denoising, magnetic field encoding enhancement, signal superposition, and relaxation correction. The final output is a corrected signal with a signal-to-noise ratio (SNR) ≥ 30 dB (based on the SNR formula SNR = 10lg(P...). signal / P noise ) Determination, P signal For signal power, P noise (Noise power), the ratio of magnetic particle signal to background signal > 50:1.

[0080] (1) Two-stage noise suppression ① First stage: Adaptive threshold filtering algorithm based on db6 wavelet basis, threshold formula Threshold=γ·sqrt(2×log(N))·P noise (γ is an adjustment factor related to magnetic field uniformity; the better the uniformity, the smaller γ becomes, to avoid over-filtering and loss of details; N is the number of signal sampling points; P) noise (Noise power), filtering out tissue-specific acoustic signals (such as low-frequency signals from vascular pulsation) and X-ray scattering noise; ② Second stage: A dedicated magnetoacoustic noise suppression module specifically filters out coil noise and pulse noise, while retaining the characteristic acoustic signals of magnetic particles in the 0-50MHz range; After the above treatment, the SNR increased from 22dB to ≥30dB (32dB in brain tracing and 33dB in wound monitoring).

[0081] (2) Magnetic field modulation signal enhancement The encoding weight ω=k×G is calculated based on the dynamic magnetic field gradient intensity G, and then the spatially variable gain formula S is used. enhanced(x,y,z,t)=Gain dynamic (x,y,z) S denoised (t) Targeted amplification of the signal in the target region, Gain dynamic Dynamically adjusted based on magnetic particle concentration and tissue depth (Gain) dynamic =Gain initial ×(C / C0)×exp (-μ×x), where C0 = 20 ng / mL is the baseline concentration, and μ = 0.05 cm⁻¹. - ¹ is the tissue attenuation coefficient); the ratio of enhanced magnetic particle signal to background is >50:1 (55:1 for brain tracing, 52:1 for lung cancer tracing).

[0082] (3) Signal superposition to enhance sensitivity The acoustic signals generated by 10 consecutive X-ray excitations and magnetic field modulations in the same region were coherently superimposed and averaged, based on the formula S. sum =∑S i (i=1 to 10), effectively improving signal sensitivity, with a detection limit of hundreds of cells (ng-level magnetic particles), and can detect target cells labeled with ≤100 magnetic particles.

[0083] (4) Relaxation effect correction Based on the magnetic particle relaxation exponential decay model h(t)=A exp (-t / T relax (A is the amplitude of the relaxation signal, T) relax (The relaxation time is calibrated for the experiment), and a deconvolution operation S is performed in the frequency domain. final (f)=S enhanced-avg (f) / H(f) (H(f) is the Fourier transform of the relaxation model), corrects for signal time-domain broadening and hysteresis caused by magnetic moment relaxation, and outputs the final purified signal S. final (t), correcting signal timing errors to ≤10ns. Where, T relax (t) A dynamic calibration mechanism is adopted: ① Based on the real-time acquisition of the magnetic field strength B by the magnetic field modulation module, the calibration is performed using formula T. relax (t) = T0 × (B0 / B) × sqrt(C0 / C) (T0 = 200 ns is the reference relaxation time, B0 = 5 mT is the reference magnetic field strength) dynamically calculated; ② After every 10 frames of imaging (1 ms), T is corrected based on the signal attenuation characteristics. relax (t), to ensure the accuracy of relaxation correction.

[0084] 3. Cell state differentiation unit Based on the attenuation coefficient characteristics of acoustic signals, magnetic particles labeled with living cells are distinguished from free magnetic particles. The threshold is calibrated through multiple sets of in vitro cell + phantom experiments, and the recognition accuracy is ≥95%.

[0085] (1) Attenuation coefficient refinement: Based on the registered magnetic particle signals, the sound wave propagation attenuation curve is calculated for each magnetic particle signal cluster, and the attenuation coefficient α is refined by the formula α=-(1 / d)lg(A / A0) (d is the propagation distance, A is the amplitude after propagation, and A0 is the initial amplitude). refined By fitting multiple frequencies from 0 to 50 MHz, the attenuation coefficient-frequency relationship α(f) is obtained, avoiding the error of single-frequency calculation.

[0086] (2) Dual-threshold three-level classification and visualization: ① Introducing tissue type adaptive threshold algorithm: Based on the acoustic impedance Z of different tissues (e.g., brain Z=1.59×10 6 Pa s / m, lung Z=0.004×10 6 Pa s / m), through the formula α th =α0×(Z / Z0) (α0=0.8dB / mm is the baseline threshold, Z0=1.5×10) 6 Pa ① The threshold is dynamically adjusted (s / m is the standard soft tissue acoustic impedance); ② Using 1MHz as the characteristic frequency, a three-level classification is performed, and the classification results are superimposed on the magnetic particle distribution image to generate a cell map with state labels: a. α refined (1MHz)≤0.8dB / mm: High-confidence viable cells (marked in dark green); b. 0.8 dB / mm < α refined (1MHz)≤1.0dB / mm: Suspected viable / transitional cells (marked in light green); c. α refined (1MHz) > 1.0dB / mm: Ionized magnetic particles (marked in yellow).

[0087] (3) Quantitative calculation of multiple parameters: ① Quantity and intensity statistics: Calculate the number (concentration) of cells / particles and the average signal intensity of each state region respectively; ② Dynamic trajectory analysis: In the time series data, track the centroid movement trajectory and signal intensity evolution of a specific cell cluster, combine the concentration distribution image and calibration coefficient k to complete the quantitative calibration of concentration, and output the correlation curve of "cell number-time" and "concentration-time".

[0088] 4. Image Reconstruction Unit By combining correction signals, ultrasonic transducer element positions, and magnetic field coding information, multi-dimensional image reconstruction is achieved through "back projection algorithm + FCS-edNET depth model + sound velocity correction", with spatial resolution ≤250μm and temporal resolution ≤100μs / frame.

[0089] (1) Basis of spatial resolution: Based on the spatial resolution formula Δs=0.8×v s / f max (v) s To organize an average sound speed of 1540 m / s, f max (assuming the ultrasonic transducer's highest receiving frequency is 5MHz), the theoretical derivation is Δs≈250μm; (2) Time resolution is controlled by the X-ray pulse frequency and the magnetic field modulation frequency. A 10kHz pulse frequency corresponds to 100μs / frame, which can capture the dynamic migration of cells in real time (such as the migration speed of stem cells at 0.2mm / h and the process of T cells gathering towards the lesion). (3) Reconstruction process and results: ① Based on the three-dimensional coordinate calculation formula (x,y,z)=f (t1,t2,...,t n )(t i (The time it takes for the signal to arrive at each ultrasonic array element), combined with a sparse reconstruction algorithm, yields the initial image; ② The localization error is ≤20μm after optimization with FCS-edNET deep model (multi-scale denoising + global-local feature fusion) and correction of propagation bias caused by tissue heterogeneity by Transformer-based sound velocity estimation algorithm; ③ The final reconstructed magnetic particle distribution image, tissue structure image, and vascular imaging image all meet the requirements of spatial resolution ≤250μm, and the vascular imaging image can identify new blood vessel branches with a diameter ≥200μm (spatial resolution ≤350μm).

[0090] 5. Image Fusion Unit Employing a pixel-level weighted fusion algorithm, it achieves precise fusion of magnetic particle distribution, tissue structure, and vascular imaging images, while also supporting cross-modal fusion with CT / MRI, with a fusion error ≤20μm.

[0091] (1) Fusion formula: F(x,y)=w1×C(x,y)+w2×T(x,y)+w3×V(x,y), where F(x,y) is the pixel value after fusion, C(x,y), T(x,y), and V(x,y) are the pixel values ​​of magnetic particle distribution, tissue structure, and blood vessel imaging images, respectively, and w1, w2, and w3 are weighting coefficients (the sum is 1, dynamically calculated based on local SNR: ω c=SNR cell / (SNR cell +SNR tissue +SNR vessel )); (2) Cross-modal fusion: Loading a pre-calibrated rigid transformation matrix T calib The reconstructed images are registered with CT / MRI data, and the initial coordinate calibration is combined with magnetic particle delivery to achieve precise matching of cell location and anatomical structure, with a fusion error of ≤20μm (18μm measured in brain tracing and 20μm measured in lung cancer tracing). (3) Data format: The fused images are automatically converted to DICOM 3.0 format, with labels for key processing parameters, compatible with hospital PACS systems, and easy for clinical review and audit tracking.

[0092] (vi) System control and display module This module is the core control hub of the system. It adopts a CPU+GPU+FPGA heterogeneous computing architecture to realize integrated control of parameters of each module, algorithm operation and result display.

[0093] 1. Hardware configuration: Industrial-grade CPU (Intel Xeon W-1370), GPU (NVIDIA A100), FPGA chip, large-capacity storage media (data transfer bandwidth ≥10Gbps), 4K high-resolution display terminal; 2. Core Functions: ① Timing control: The FPGA generates nanosecond-level synchronous trigger signals to ensure that all modules work together. The synchronization error is ≤10ns. In case of abnormal conditions (excessive radiation dose, abnormal magnetic field uniformity), the scanning will be stopped automatically and an alarm will be triggered. ② Algorithm operation: GPU-accelerated image reconstruction and fusion algorithm, reconstruction time <1s / frame, meeting the needs of real-time clinical monitoring; ③ Results display and storage: The 4K terminal synchronously displays the fused image, signal intensity-time curve, and cell number-time curve, supports image scaling and measurement (cell number, migration distance, blood vessel area), and the raw data is preserved without loss. The processed results are exported in DICOM format. ④ Human-computer interaction: Provides dedicated parameter packages for multiple scenarios (stem cell tracing, CAR-T monitoring, wound repair, etc.), supports custom parameter adjustment, visualizes the processing chain, and allows real-time viewing of signal and image results at each step.

[0094] To illustrate the solution of this embodiment more clearly, a specific scenario is used as an example.

[0095] Example 1: For a patient undergoing dopaminergic stem cell transplantation, the dopaminergic stem cells first need to be magnetically labeled.

[0096] Preparation phase: 1. Magnetic particle labeling: SPIONs (with a particle size controlled at around 15 nm) with a surface modified with CD133 antibody (a stem cell-specific marker) are labeled with dopaminergic stem cells (1×10⁻⁶). 6 The cells were cultured on an order of magnitude (likely referring to a specific order of magnitude), and the labeling rate was dynamically detected by flow cytometry. A labeling rate of 96% was considered acceptable. It should be noted that the SPIONs used in this embodiment have passed the ISO 10993 biocompatibility test. Acute toxicity tests showed an LD50 > 500 mg / kg (mouse model), and long-term metabolic tests showed that more than 90% of the magnetic particles could be metabolized and excreted from the body through the liver, with no risk of accumulation in the body.

[0097] 2. Cell Delivery: Using stereotactic craniotomy, labeled stem cells are precisely delivered to the substantia nigra region of the patient's brain via a micro-injector (e.g., 10 μL) (coordinates: 3 mm to the right of the midpoint of the anterior commissure-posterior commissure line, 5 mm below, and 12 mm deep). The optical positioning system is calibrated in real time, and the final delivery error is ≤30 μm. Postoperative CT confirms no bleeding or tissue damage. The delivery coordinates (x0=45 cm, y0=30 cm, z0=75 cm) are uploaded to the control host for module parameter calibration.

[0098] Imaging stage: 1. System parameter calibration: The main controller loads the "brain stem cell tracing" parameters.

[0099] The X-ray signal transmitting module receives commands from the control host, sets the tube voltage to 80kVp (corresponding to X-ray energy of 30keV), pulse width to 10ns, pulse frequency to 10kHz (corresponding to time resolution of 100μs / frame), and the collimator controls the beam diameter to 2mm, focusing on the substantia nigra region of the brain; the dose monitoring unit reports that the radiation dose per scan is 0.08mSv, which meets the preset requirements. Magnetic field modulation module: Receives synchronous trigger signals and control host commands, sets the static magnetic field strength to 5mT, the dynamic gradient magnetic field to 0.3mT / m (coordinated modulation in x / y / z directions), the frequency to 30kHz, and the magnetic field distribution to cover the substantia nigra region of the brain and the surrounding 2cm range, with uniformity <5%. During signal reception, the control host adjusts the center frequency of the ultrasound transducer to 20MHz, the focusing depth to 12mm, and the receiving gain to 45dB according to the delivery coordinates. The robotic arm adjusts its angle to 30° (avoiding the strong reflection area of ​​the skull). The main controller generates a synchronization clock to ensure that the timing of each X-ray pulse emission, the timing of the gradient field switching to the corresponding coded value, and the timing of the opening of the ultrasound acquisition window are strictly aligned, with a synchronization error of ≤8ns.

[0100] 2. Dynamic scanning and data acquisition: After the system is started, a "layered scanning + dynamic tracking" mode is adopted: first, a three-dimensional layered scan (50μm layer thickness) is performed on the substantia nigra region of the brain and the surrounding 2cm range to obtain an initial cell distribution image; then, a continuous dynamic scan is performed for 30 minutes, storing a complete frame of image data (including cell distribution, tissue structure, signal attenuation coefficient, and magnetic field modulation parameters) every 10 seconds, with a total data volume of approximately 1.2GB. This data is transmitted to the control host in real time via Ethernet with suitable bandwidth. It should be noted that the image data format of this embodiment is compatible with the DICOM 3.0 standard and can be seamlessly integrated with the hospital's existing PACS system, facilitating clinical doctors' access and archiving.

[0101] 3. Signal Processing and Image Analysis: Signal denoising: The acquired acoustic signal was subjected to 5-layer threshold filtering using the db6 wavelet basis to remove skull reflection noise (frequency < 1MHz) and vascular pulsation interference (frequency 1-2Hz). The signal-to-noise ratio was improved to 32dB after filtering.

[0102] Magnetic field modulation signal enhancement: Based on the dynamic magnetic field gradient intensity of 0.3mT / m, the coding weight ω=0.05×0.3=0.015 is calculated to specifically amplify the signal in the substantia nigra region of the brain, suppress background interference from brain tissue, and improve the signal specificity to 55:1.

[0103] Relaxation effect correction: The signal lag caused by magnetic particle relaxation is corrected by deconvolution algorithm, and the time resolution is further optimized to 150μs / frame.

[0104] Survival status determination: The signal attenuation coefficient α is extracted and identified by a threshold algorithm: regions with α ≤ 0.7 dB / mm are identified as surviving stem cells, and regions with α > 1.1 dB / mm are identified as free magnetic particles.

[0105] Quantitative analysis: The number of surviving stem cells was counted using image analysis software, the migration distance was calculated (with the initial delivery location as the origin), and curves of "cell number - time" and "migration distance - time" were generated.

[0106] Implementation Results Analysis Phase: 1. Imaging performance indicators: Spatial resolution: The smallest identifiable unit of the image of stem cell distribution in the substantia nigra region of the brain is 250 μm, which is consistent with the calculation result of Δs=0.8×1540m / s / 5MHz≈250μm. It clearly shows three independent stem cell aggregation regions (containing 230, 180, and 150 cells respectively, corresponding to magnetic particle amounts of 11.5ng, 9ng, and 7.5ng), which is superior to MRI (1.5mm), ultrasound (200μm) and traditional XMPAI (100μm).

[0107] Temporal resolution: Dynamic imaging at 100 μs / frame can capture the slow migration process of stem cells (average migration speed 0.2 mm / h). At 30 minutes, some stem cells were observed to move towards the globus pallidus region (migration distance 0.8 mm), providing dynamic evidence for subsequent efficacy evaluation.

[0108] Sensitivity: Successfully detected 100 isolated surviving stem cells in the white matter region of the brain (corresponding to 5ng of magnetic particles), proving that the system can achieve sensitive detection at a small cell scale, thanks to the signal amplification effect of magnetic field modulation.

[0109] 2. Clinical value verification: Follow-up imaging at 7, 14, and 28 days post-surgery showed that the number of surviving stem cells remained at 75%-80% of the initial value (with no obvious apoptosis) and gradually migrated to the damaged dopaminergic neurons (confirmed by preoperative PET), consistent with the trend of relief of the patient's left limb tremor symptoms (tremor score decreased by 35% at 28 days post-surgery), providing accurate data for the evaluation of treatment effect.

[0110] No significant aggregation of free magnetic particles was detected (the area marked in yellow accounts for less than 5%), proving that the stem cells are in good condition and there is no large release of magnetic particles. This avoids the problem of misjudging "cell survival" by traditional MRI and reduces unnecessary secondary treatment.

[0111] In another scenario, this embodiment uses CD19-targeted CAR-T cell therapy on a non-small cell lung cancer patient with three lesions in the right lobe of the lung who has failed chemotherapy.

[0112] Preparation phase: 1. Magnetic particle labeling: SPIONs (particle size controlled at approximately 12 nm, magnetic saturation intensity at approximately 75 emu / g) with surface-modified CD3 antibody (T cell-specific marker) are labeled with CAR-T cells (5 × 10⁻⁶ cells / g). 7The cells were co-cultured at 37°C and 5% CO2 for 16 hours. When the labeling rate reached 98%, it was considered to meet the requirements. The magnetic particles were observed by transmission electron microscopy to confirm that they were evenly distributed in the cytoplasm of T cells and had no obvious cytotoxicity (viability > 95%). 2. Cell delivery: Labeled CAR-T cells are infused via peripheral vein at an infusion rate of 5 × 10⁻⁶. 6 The infusion rate is 1 unit / min, and the first imaging monitoring is initiated 24 hours after infusion.

[0113] Imaging stage: 1. System parameter calibration: The main controller loads the "CAR-T cell tracing" parameters. The X-ray signal transmitting module receives instructions from the control host and sets the tube voltage to 150kVp (corresponding to X-ray energy of 60keV), pulse width to 10ns, pulse frequency to 10kHz (time resolution 100μs / frame), and single scan radiation dose to 0.09mSv.

[0114] Magnetic field modulation module: Receives synchronous trigger signals and control host commands, sets the static magnetic field strength to 8mT, the dynamic gradient magnetic field to 0.4mT / m, and the frequency to 40kHz, performs magnetic field spatial encoding on the lung lesion area, and suppresses lung gas scattering interference.

[0115] During signal reception, the control host adjusts the center frequency to 20MHz (to reduce the influence of lung gas scattering) and the focusing depth to 8cm (aiming at the lesion area in the right lung lobe) according to the delivery coordinates, and adopts a multi-element synchronous reception mode (128 elements working simultaneously) to improve signal acquisition efficiency.

[0116] 2. Dynamic scanning and data acquisition: The "dynamic region tracking" algorithm is enabled to automatically lock the lung lesion area (by preoperative enhanced CT positioning). Combined with the spatial coding information of magnetic field modulation, only the lesion and the surrounding 3cm range are scanned to reduce the amount of invalid data.

[0117] 3. Test Results and Analysis: 24 hours after infusion: CAR-T cell aggregation was detected in all three lesion areas of the right lung lobe, with approximately 3 × 10⁻⁶ cells per lesion area. 5 The signal attenuation coefficient α was 0.6 dB / mm (indicating surviving T cells), and magnetic field modulation increased the signal intensity by 4 times, proving that the T cells successfully migrated to the lesion.

[0118] Seven days after infusion: The number of T cells in the lesion area increased to 1.8 times the initial value and aggregated towards the center of the lesion (in contact with tumor cells). Simultaneously acquired tissue structure images showed necrotic areas at the edge of the lesion (grayscale value decreased by 20%).

[0119] 21 days after infusion: the largest lesion diameter shrank to 1.2 cm, the number of T cells remained at a high level (1.5 times the initial value), and no obvious free magnetic particles were detected, proving that CAR-T cells continued to exert anti-tumor effects.

[0120] Example 2: For a patient receiving CD19-targeted CAR-T cell therapy.

[0121] ABS nanocomposites (particle size 12 nm, magnetic saturation strength 75 emu / g, Bi:Fe=5:1) with surface-modified CD3 antibody (T cell specific marker) were co-cultured with CAR-T cells (5×107 cells) at 37℃ and 5% CO2 for 16 hours. The labeling efficiency reached 98%, and the cell viability was verified to be >95% by MTT assay. Transmission electron microscopy confirmed that the magnetic particles were uniformly distributed in the cytoplasm of T cells without obvious aggregation. (3) Cell infusion: Labeled CAR-T cells were infused via a peripheral vein at an infusion rate of 5 × 10⁻⁶. 6 The infusion rate was 1 unit / min, and the first imaging monitoring was initiated 24 hours after infusion; preoperative CT data was imported for cross-modal fusion.

[0122] 2. Imaging process (1) Parameter settings: Focusing and positioning unit: Receives the coordinates of 3 lesions, drives the X-ray generator to move sequentially and completes laser positioning (positioning deviation ≤90μm each time).

[0123] Ultrashort pulse X-ray: tube voltage 150kVp (corresponding to X-ray energy 60keV), pulse width 10ns, pulse frequency 10kHz (time resolution 100μs / frame), tube current 1mA, scan time 9s, sparse reconstruction algorithm enabled, single scan radiation dose 0.09mSv.

[0124] Precision magnetic field modulation module: static magnetic field strength 8mT, dynamic gradient magnetic field 0.4mT / m, frequency 40kHz, adapted to the magnetic particles in this embodiment to ensure high magnetic particle signal response intensity; performs magnetic field spatial encoding on the lung lesion area to suppress lung gas scattering interference; monitors magnetic field uniformity in real time and feeds back data to ensure uniformity <5%.

[0125] Acoustic signal detection module: The center frequency of the ultrasound transducer is 20MHz (to reduce the influence of gas scattering in the lungs), the focusing depth is 8cm (aiming at the lesion area in the right lung lobe), and a 32-element synchronous receiving mode is adopted to improve signal acquisition efficiency.

[0126] Image reconstruction: The "dynamic region tracking + sparse reconstruction" algorithm is enabled to automatically locate the lung lesion area (by preoperative enhanced CT positioning). Combined with the spatial coding information of magnetic field modulation, only the lesion and the surrounding 3cm range are scanned, reducing the amount of invalid data by 60% and improving reconstruction efficiency.

[0127] (2) Test results and analysis: 24 hours post-infusion: CAR-T cell aggregation was detected in all three lesion areas of the right lung lobe (marked in dark green), with approximately 3 × 10⁻⁶ cells per lesion area. 5 The signal attenuation coefficient α was 0.6 dB / mm (indicating surviving T cells), and magnetic field modulation increased the signal intensity by 4 times, proving that the T cells successfully migrated to the lesion.

[0128] Seven days after infusion: The number of T cells in the lesion area increased to 1.8 times the initial value and aggregated towards the center of the lesion (in contact with tumor cells). Simultaneously acquired tissue structure images showed necrotic areas at the edge of the lesion (grayscale value decreased by 20%).

[0129] 21 days after infusion: the largest lesion diameter shrank to 1.2 cm, the number of T cells remained at a high level (1.5 times the initial value), and no obvious free magnetic particles were detected, proving that CAR-T cells continued to exert anti-tumor effects.

[0130] 3. Implementation Results (1) Technical Results: The algorithm runs stably without data loss or processing anomalies. During the 21-day follow-up imaging, the tracking error of the deep learning model for low-concentration T cells in the lesion area was ≤15μm with no obvious artifacts, and the accuracy of lesion diameter measurement was consistent with the pathological results at 97%. The reconstructed images clearly showed the aggregation and proliferation of CAR-T cells in the lesion area. The maximum lesion diameter decreased from 2.5cm to 1.2cm within 21 days, which was negatively correlated with the change in T cell count. With a spatial resolution of 250μm and a temporal resolution of 100μs / frame, accurate concentration quantification was achieved, and the reconstruction time was 0.9s / frame, meeting the accuracy and real-time requirements of clinical monitoring.

[0131] (2) Clinical effect: Accurately track the migration, aggregation and survival status of CAR-T cells, providing doctors with direct evidence to judge the effectiveness of treatment, avoiding blindly continuing treatment or changing the plan, and reducing the medical costs of patients; cross-modal navigation map can assist in the adjustment of subsequent treatment plans and improve the accuracy of diagnosis and treatment.

[0132] Example 3: Monitoring angiogenesis during wound repair in experimental rats.

[0133] (1) Experimental subjects: A full-thickness skin wound model of the back of SD rats (wound diameter 1.5cm, depth to the subcutaneous fascia layer) was constructed. A total of 12 rats were randomly divided into an experimental group (receiving angiogenesis promotion therapy) and a control group (routine care). Imaging monitoring was started 24 hours after modeling. (2) Magnetic particle delivery: ABS nanocomposite with surface-modified CD31 antibody (a specific marker for vascular endothelial cells) (particle size 18 nm, Bi:Fe = 5:1, saturation magnetization 72 emu / g) was injected subcutaneously around the wound (4 injection points, 50 μL per point, magnetic particle concentration 20 ng / μL) to ensure that the magnetic particles targeted and bound to newly formed vascular endothelial cells; MTTassay verified that the effect of magnetic particles on endothelial cell viability was >95%, and flow cytometry detected a targeting binding efficiency ≥92%; (3) Preoperative localization: Two-dimensional images of the wound area on the back of the rat were acquired by ultrasound diagnostic equipment. The center of the wound and the surrounding 2cm range were determined as the region of interest. The three-dimensional coordinates (x0=35cm, y0=25cm, z0=60cm) were extracted and uploaded to the control host. The preoperative mini CT angiography data were imported for tissue structure reference and cross-modal fusion.

[0134] 2. Imaging process (1) System parameter calibration: The main controller loads a dedicated parameter package for "wound microvascular imaging" to adapt to the imaging needs of superficial tissues and blood vessels.

[0135] Target area focusing and positioning module: Receives the coordinates of the region of interest, drives a three-way stepper motor to adjust the position of the X-ray generator (vertical distance of 60cm on the Z-axis), uses laser positioning combined with Hall magnetic field sensor array calibration, and finally achieves an alignment deviation of ≤75μm between the X-ray focal point, magnetic field modulation area, and ultrasonic focal point, and outputs a "positioning ready" signal.

[0136] Ultrashort pulse X-ray module: tube voltage set to 120kVp (corresponding to X-ray energy of 45keV), pulse width 15ns, pulse frequency 10kHz (time resolution 100μs / frame), collimator controls beam diameter of 3mm to cover the entire wound area; tube current 1mA, scan time 7s, sparse reconstruction algorithm enabled, dose monitoring unit reports single scan radiation dose of 0.07mSv, which meets animal experiment safety standards.

[0137] Precision magnetic field modulation module: static magnetic field strength 6mT, dynamic gradient magnetic field 0.35mT / m (mainly in the x / y direction, adapting to the planar distribution of the wound), frequency 35kHz, magnetic field uniformity <5%; optimized magnetic field spatial coding weights to suppress background noise from subcutaneous adipose tissue, taking into account the loose characteristics of wound tissue.

[0138] Acoustic signal detection module: Adjust the center frequency of the ultrasound transducer to 15MHz (to balance the penetration depth and vascular resolution), focus depth to 5mm (to align with the subcutaneous vascular layer of the wound), receive gain to 40dB, adjust the robotic arm angle to 45° to reduce interference from scar tissue reflection on the wound surface; enable multi-element synchronous receiving mode to improve the efficiency of microvascular signal acquisition.

[0139] Synchronous triggering: The FPGA generates a synchronous clock to ensure that the synchronization error of X-ray pulse emission, magnetic field gradient switching and ultrasound acquisition window opening is ≤9ns, which is suitable for the rapid capture of dynamic blood flow in blood vessels.

[0140] (2) Dynamic scanning and data acquisition: The study adopted a "daily static scanning + weekly dynamic tracking" model: static 3D scanning (40μm slice thickness) was performed on days 1, 3, 5, 7 and 14 after modeling to obtain images of vascular distribution and magnetic particle binding; an additional 1-hour dynamic scan was performed on day 7 (one frame was stored every 30 seconds) to monitor changes in vascular hemodynamics; all data were transmitted to the control host in real time via 10Gbps Ethernet, with a single frame data size of approximately 800MB.

[0141] Wound healing indices (wound contraction rate and epithelialization rate) of the two groups of rats were recorded simultaneously for subsequent correlation analysis between imaging results and clinical efficacy.

[0142] (3) Signal processing and image reconstruction: Magnetic field encoded signal enhancement: ① Two-stage denoising: The first stage uses db6 wavelet basis filtering to remove wound tissue motion noise (frequency 0.5-1Hz), and the second stage uses a dedicated magnetoacoustic noise suppression module to filter out subcutaneous fat scattering noise, improving the signal-to-noise ratio to 33dB after filtering; ② Magnetic field encoded weighting: Based on the dynamic gradient magnetic field strength of 0.35mT / m, the encoding weight ω=0.05×0.35=0.0175 is calculated to specifically amplify the magnetic particle signal in the vascular region, making the ratio of magnetic particle signal to background reach 53:1; ③ Relaxation correction: By dynamically calibrating the relaxation time Trelax=180ns (adapting to the wound tissue environment), the deconvolution algorithm corrects the signal lag, reducing the timing error to ≤9ns.

[0143] Spatiotemporal collaborative image reconstruction: ① Signal feature extraction: The propagation time ti and amplitude A of the ultrasound array element received signal are extracted. Combined with the magnetic particle targeting binding characteristics, the three-dimensional coordinates of microvessels are restored through back projection algorithm and sparse reconstruction algorithm. The initial reconstructed image spatial resolution is 280μm; ② FCS-edNET depth optimization: The multi-scale denoising prior module suppresses wound edge artifacts, and the MSGL module enhances the fine branch features of microvessels. After optimization, the spatial resolution is improved to 250μm; The Transformer-based sound velocity estimation algorithm corrects the sound velocity differences in subcutaneous tissue (fat, fascia), and the positioning error is ≤18μm; ③ Microvessel imaging branch processing: Utilizing the ultrasound microbubble resonance effect (ultrasound length λ=3μm, matching microbubble diameter 3μm), the magnetic particle-microbubble composite probe signal in the blood vessel is amplified to generate a three-dimensional image of microvessels with a spatial resolution ≤350μm. It can identify new blood vessel branches with a diameter ≥200μm.

[0144] Cell / Vascular State Recognition and Image Fusion: ① State Differentiation: The acoustic signal attenuation coefficient α of magnetic particles is extracted, and the dynamic threshold is adjusted based on the characteristics of wound tissue: α≤0.78dB / mm indicates magnetic particles that target and bind to neovascularization (marked in dark green), 0.78dB / mm<α≤0.98dB / mm indicates free magnetic particles (marked in light green), and α>0.98dB / mm indicates non-specifically bound magnetic particles (marked in yellow), with a recognition accuracy of 95.3%. ② Pixel-Level Fusion: A pre-calibrated rigid transformation matrix Tcalib is loaded to fuse microvascular images with preoperative CT data and wound ultrasound images, with a fusion error ≤19μm; parameters such as microvascular density (MVD), vascular branch length, and neovascularization area ratio are quantitatively extracted.

[0145] 3. Implementation Results (1) Technical effects: Imaging performance: Spatial resolution of 250μm and temporal resolution of 100μs / frame, which can clearly present the three-dimensional distribution and branching morphology of neovascularization around the wound; Sensitivity of 8ng-level magnetic particle detection, which can capture magnetic particle signals around a single neovascular branch (200μm in diameter).

[0146] Dynamic monitoring capability: The microvascular density in the experimental group reached 28.6 vessels / mm² 7 days after surgery, which was 75.4% higher than that in the control group (16.3 vessels / mm²). Dynamic scanning showed that the blood flow signal intensity in the experimental group increased linearly over time (slope 0.03dB / s), while the increase in the control group was slow (slope 0.012dB / s), which was highly correlated with the rate of angiogenesis.

[0147] System stability: No issues such as excessive radiation or magnetic field drift were observed during 14 consecutive days of monitoring. The radiation dose per scan was 0.07 mSv, the cumulative radiation dose was <1 mSv, and there was no significant tissue damage.

[0148] (2) Clinical value verification: Precise assessment of treatment efficacy: By using quantitative parameters such as microvessel density and branch length, the differences in angiogenesis between the experimental group and the control group can be objectively distinguished, providing a visual basis for evaluating the efficacy of angiogenesis-promoting drugs.

[0149] Treatment guidance and adjustment: Monitoring of two control group rats revealed delayed angiogenesis (MVD < 12 lines / mm² 7 days post-surgery). After timely adjustment of the treatment plan, the wound contraction rate increased from 42% to 68% 14 days later, verifying the clinical guidance value of the technology.

[0150] Application adaptability: The device is small in size (4.5m³) and can be used for monitoring small wounds (such as diabetic foot ulcers and burn wounds) in animal experiments and clinical settings. The operation process is simplified and medical staff can complete the imaging independently after 1 hour of training.

[0151] In summary, the solution presented in this embodiment provides precise cell tracing data for stem cell therapy of diseases such as Parkinson's disease and spinal cord injury, as well as for CAR-T therapy for tumors. This helps doctors assess treatment effectiveness in a timely manner, adjust treatment plans, reduce ineffective treatments, and improve treatment success rates. Secondly, the low radiation dose reduces the health risks of long-term monitoring, eliminates the need for a dedicated shielded room, and simplifies the examination process. Real-time imaging reduces patient waiting time, avoids multiple scans, and improves the patient experience. In addition to stem cell tracing and tumor T-cell therapy monitoring, the solution can be extended to scenarios such as "magnetically labeled mesenchymal stem cells for bone repair monitoring," "magnetically labeled dendritic cells for immunotherapy evaluation," and "microvascular imaging," covering core areas of regenerative medicine and tumor treatment, and has broad clinical application value.

[0152] Based on the ultrashort pulse X-ray excited magnetic particle acoustic imaging method of the above embodiments, this application also provides an ultrashort pulse X-ray excited magnetic particle acoustic imaging device, such as... Figure 3 As shown, Figure 3 This is a schematic diagram of the structure of an ultrashort pulse X-ray excited magnetic particle acoustic imaging device provided in an embodiment of this application. The ultrashort pulse X-ray excited magnetic particle acoustic imaging device 3 includes a processor 301 and a memory 302. The memory 302 is used to store computer programs; the processor 301 is used to call and run the computer programs from the memory to execute the ultrashort pulse X-ray excited magnetic particle acoustic imaging method as described in the above embodiment.

[0153] In the embodiments of this application, the processor 301 described above can be at least one of the following: Application-Specific Integrated Circuit (ASIC), Digital Signal Processor (DSP), Digital Signal Processing Device (DSPD), Programmable Logic Device (PLD), Field Programmable Gate Array (FPGA), Central Processing Unit (CPU), Controller, Microcontroller, and Microprocessor. It is understood that for different devices, the electronic device used to implement the above processor function can also be other types, and the embodiments of this application do not specifically limit it.

[0154] This application provides a computer-readable storage medium storing a computer program for implementing, when executed by a processor, the magnetic particle acoustic imaging method excited by ultrashort pulse X-rays as described in any of the above embodiments.

[0155] For example, the program instructions corresponding to the ultrashort pulse X-ray excited magnetic particle acoustic imaging method in this embodiment can be stored on storage media such as optical discs, hard disks, and USB flash drives. When the program instructions corresponding to the ultrashort pulse X-ray excited magnetic particle acoustic imaging method in the storage medium are read or executed by an electronic device, the ultrashort pulse X-ray excited magnetic particle acoustic imaging method as described in any of the above embodiments can be realized.

[0156] Furthermore, the functional modules in the embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional module.

[0157] If the integrated unit is implemented as a software functional module and is not sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this embodiment, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the method of this embodiment. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0158] It should be understood that the phrases "one embodiment," "an embodiment," or "some embodiments" mentioned throughout the specification mean that a specific feature, structure, or characteristic related to an embodiment is included in at least one embodiment of this application. Therefore, "in one embodiment," "in one embodiment," or "in some embodiments" appearing throughout the specification do not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. It should be understood that in the various embodiments of this application, the sequence numbers of the above-described processes do not imply a sequential order of execution; the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application. The sequence numbers of the above-described embodiments are merely for descriptive purposes and do not represent the superiority or inferiority of the embodiments. The descriptions of the various embodiments above tend to emphasize the differences between the various embodiments; their similarities or commonalities can be referred to mutually, and for the sake of brevity, they will not be repeated here.

[0159] The modules described above as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules. They may be located in one place or distributed across multiple network units. Some or all of the modules may be selected to achieve the purpose of this embodiment according to actual needs.

[0160] In addition, each functional module in the various embodiments of this application can be integrated into one processing unit, or each module can be a separate unit, or two or more modules can be integrated into one unit; the integrated modules can be implemented in hardware or in the form of hardware plus software functional units.

[0161] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media that can store program code, such as mobile storage devices, read-only memory (ROM), magnetic disks, or optical disks.

[0162] The methods disclosed in the several method embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments.

[0163] The features disclosed in the several product embodiments provided in this application can be arbitrarily combined without conflict to obtain new product embodiments.

[0164] The features disclosed in the several method or device embodiments provided in this application can be arbitrarily combined without conflict to obtain new method or device embodiments.

[0165] The above description is merely an embodiment of this application, but the protection scope of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the protection scope of this application. Therefore, the protection scope of this application should be determined by the protection scope of the claims.

Claims

1. A magnetic particle acoustic imaging system excited by ultrashort pulse X-rays, characterized in that, include: An X-ray signal emission module is used to respond to control commands and control an X-ray generator to generate X-ray pulses with preset parameters, so that the X-ray pulses act on a pre-set magnetic particle marking region, wherein the magnetic particles include a magnetic core, a functional modification layer and a biocompatible matrix, and the magnetic particle ligands include targeting molecules. A magnetic field modulation module is used to modulate the magnetic particle marking area so that when the X-ray pulse acts on the magnetic particles, it generates an original acoustic signal that matches the magnetic field modulation. An acoustic detection module is used to convert the raw acoustic signal into a raw data signal; The signal processing module is used to receive the raw data signal and process the raw data signal to obtain the imaging result.

2. The magnetic particle acoustic imaging system excited by ultrashort pulse X-rays according to claim 1, characterized in that, The X-ray signal transmitting module includes: A position control unit is used to receive external feedback information and adjust the X-ray generator to the corresponding position according to the external feedback information. The signal transmitting unit is used to control the X-ray generator to generate an X-ray pulse with preset parameters when the corresponding position coincides with the preset magnetic particle marking area, wherein the preset parameters include tube voltage, pulse width, and pulse frequency.

3. The magnetic particle acoustic imaging system excited by ultrashort pulse X-rays according to claim 1, characterized in that, The magnetic field modulation module includes: A static magnetic field modulation unit is used to generate a magnetic field signal of fixed intensity to magnetize magnetic particles, so that the magnetic moments of the magnetic particles are aligned in a direction that matches the magnetic field signal of fixed intensity. A dynamic magnetic field modulation unit is used to generate a gradient magnetic field signal based on the fixed-intensity magnetic field signal so that the fixed-intensity magnetic field signal and the gradient magnetic field signal form a composite magnetic field signal for encoding the composite signal.

4. The magnetic particle acoustic imaging system excited by ultrashort pulse X-rays according to claim 3, characterized in that, The dynamic magnetic field modulation unit includes: A waveform generator is used to generate corresponding alternating signals according to frequency commands. A gradient amplifier is used to output an adjustable drive current signal according to the alternating signal, so as to generate a corresponding gradient magnetic field signal through a gradient coil.

5. The magnetic particle acoustic imaging system excited by ultrashort pulse X-rays according to claim 3, characterized in that, It also includes a signal synchronization module, which, in response to the X-ray signal transmitting module sending an X-ray pulse, sends a synchronization trigger signal to the magnetic field modulation module and the acoustic detection module so that the signal frequencies of the X-ray signal transmitting module and the magnetic field modulation module are matched.

6. The magnetic particle acoustic imaging system excited by ultrashort pulse X-rays according to claim 1, characterized in that, The signal processing module includes: The acoustic signal extraction unit is used to acquire raw data signals and process them to obtain relaxation correction signals. An image reconstruction unit is used to convert the relaxation correction signal into a magnetic particle distribution image and a concentration distribution image; The cell state recognition unit is used to determine the attenuation threshold classification based on the acoustic impedance of different tissues to determine the cell state based on the attenuation threshold classification, to label the cell state to the magnetic particle distribution image to generate a cell map with state labels, and to determine quantitative concentration data based on the concentration distribution image and the cell state. An image fusion unit is used to fuse a cell image with status labels, concentration quantification data, and an anatomical structure image corresponding to the cell image with status labels to obtain a fused image, wherein the anatomical structure image includes a tissue structure image and a microvascular imaging image.

7. The magnetic particle acoustic imaging system excited by ultrashort pulse X-rays according to claim 6, characterized in that, The acoustic signal extraction unit includes: A noise suppression subunit is used to filter the original acoustic image according to a predetermined filtering threshold to obtain a denoised signal, wherein the filtering threshold is determined based on the noise power within the signal frequency band; A signal enhancement subunit is used to perform signal gain on the denoised signal to amplify the signal intensity in the high gradient region to obtain an enhanced denoised signal. The signal averaging and sensitivity enhancement subunit is used to coherently superimpose and average the amplified enhanced and denoised signal, and then filter the effective data to obtain the average sensitivity enhancement signal. The relaxation correction subunit is used to deconvolve the average sensitivity enhancement signal in the frequency domain to obtain the relaxation correction signal.

8. A method for acoustic imaging of magnetic particles excited by ultrashort pulse X-rays, characterized in that, include: In response to a control command, the X-ray generator is controlled to generate an X-ray pulse with preset parameters, such that the X-ray pulse acts on a pre-set magnetic particle-marked region, wherein the magnetic particles include a magnetic core, a functional modification layer and a biocompatible matrix, and the magnetic particle ligands include a targeting molecule. The magnetic particle-marked region is modulated with a magnetic field so that when the X-ray pulse acts on the magnetic particles, it generates an original acoustic signal that matches the magnetic field modulation. The original acoustic signal is converted into a raw data signal, and the raw data signal is processed to obtain the imaging result.

9. The method for magnetic particle acoustic imaging excited by ultrashort pulse X-rays according to claim 8, characterized in that, The raw data signal is processed to obtain the imaging result, including: The raw data signal is acquired and processed to obtain the relaxation correction signal; The relaxation correction signal is converted into a magnetic particle distribution image and a concentration distribution image; The attenuation threshold is determined based on the acoustic impedance of different tissues to classify cell states, and the cell states are labeled onto the magnetic particle distribution image to generate a cell map with state labels. Concentration quantitative data is determined based on the concentration distribution image and the cell states. A fused image is obtained by fusing a state-labeled cell map, quantitative concentration data, and an anatomical structure map corresponding to the state-labeled cell map, wherein the anatomical structure map includes a tissue structure map and a microvascular imaging map.

10. The magnetic particle acoustic imaging system excited by ultrashort pulse X-rays according to claim 9, characterized in that, The raw data signals are acquired and processed to obtain relaxation correction signals, including: The original acoustic image is filtered according to a predetermined filtering threshold to obtain a denoised signal, wherein the filtering threshold is determined based on the noise power within the signal frequency band; The denoised signal is amplified by signal gain to increase the signal intensity in the high gradient region, thereby obtaining an enhanced denoised signal. The amplified and denoised signal is coherently superimposed and averaged, and then the effective data is filtered to obtain the average sensitivity-enhancing signal. The relaxation correction signal is obtained by deconvolving the average sensitivity enhancement signal in the frequency domain.

Images