Magnetic nanoparticle imaging fundamental frequency signal compensation method based on FPGA

Abstract

The invention belongs to the technical field of magnetic nanoparticle imaging, and particularly relates to a magnetic nanoparticle imaging fundamental frequency signal compensation method based on an FPGA (Field Programmable Gate Array), which comprises the following steps: S1, sending an excitation signal to a power amplifier by the FPGA; the frequency of the excitation signal is between 1kHz and 100kHz; and S2, transmitting the signal to a transmitting coil by a power amplifier, generating an excitation field by the transmitting coil, receiving the signal by a first-order gradient meter, and carrying out passive compensation on the received feed-through signal. A feed-through signal is accurately counteracted through FPGA real-time calculation and inversion signal modulation, the high-efficiency calculation capability of the FPGA supports real-time signal processing and adapts to the change of an excitation magnetic field, the fundamental component of a particle signal is recovered through an iterative compensation process, the signal integrity is ensured, the hardware complexity is reduced through a digital method, and the calculation efficiency is improved. And the stability and repeatability of the system are improved.

Description

Technical Field

[0001] The invention relates to the technical field of magnetic nanoparticle imaging, in particular to a magnetic nanoparticle imaging fundamental frequency signal compensation method based on FPGA. Background Art

[0002] MPI is a new type of medical imaging technology that uses superparamagnetic iron oxide nanoparticles (SPIONs) as targeted tracers to measure the concentration distribution of magnetic nanoparticles through their nonlinear response characteristics in the excitation field. MPI has the advantages of high sensitivity, high spatial resolution, high temporal resolution and no radiation, and is widely used in biomedical fields such as tumor imaging, neuroimaging, vascular imaging and perfusion imaging.

[0003] In the MPI system, the excitation magnetic field will directly couple to the receiving coil, forming a strong feedthrough signal, which is several orders of magnitude higher than the particle signal, seriously interfering with the detection of SPIONs signals. In the prior art, the methods to eliminate the feedthrough signal mainly include filtering and passive compensation, but these methods have problems such as insufficient compensation accuracy, loss of fundamental frequency signal, poor real-time dynamic adjustment capability, and insufficient system complexity and stability. Therefore, we proposed a FPGA-based magnetic nanoparticle imaging fundamental frequency signal compensation method to solve the above problems. Summary of the invention

[0004] 1. Technical issues to be resolved

[0005] In view of the deficiencies in the prior art, the present invention provides a method for compensating fundamental frequency signals of magnetic nanoparticle imaging based on FPGA, which solves the problems raised in the above-mentioned background technology.

[0006] (II) Technical solution

[0007] In order to achieve the above-mentioned purpose, the present invention specifically adopts the following technical solutions:

[0008] A method for compensating a fundamental frequency signal of magnetic nanoparticle imaging based on FPGA comprises the following steps:

[0009] S1: The field programmable gate array (FPGA) sends an excitation signal to the power amplifier; the excitation signal frequency is between 1kHz and 100kHz;

[0010] S2: The power amplifier then transmits the signal to the transmitting coil, which generates an excitation field. The first-order gradiometer receives the signal and passively compensates the received feedthrough signal.

[0011] S3: The passively compensated feedthrough signal is first converted into a digital signal through an analog-to-digital converter (ADC), and then the digital signal is transmitted to the FPGA, and the amplitude and phase of the feedthrough signal are further calculated using a dynamic feedthrough compensation algorithm;

[0012] S4: The FPGA generates a modulation signal with a phase opposite to that of the feedthrough signal and transmits the signal to a digital-to-analog converter (DAC). The DAC then sends the converted analog modulation signal to a differential low-noise amplifier.

[0013] S5: Repeat steps S3 and S4 until the strength of the acquired baseband signal is equivalent to the noise floor level, which indicates that the compensation effect has achieved the expected goal;

[0014] S6: Adjust the gain of the low noise amplifier. The feedthrough signal is first amplified by the A channel of the low noise amplifier and then subtracted from the modulated signal. The difference signal is sent to the phase-locked amplifier to calculate the strength of the baseband signal. If the calculated baseband signal strength is not equivalent to the background noise level, then the calculation result is divided by the gain, and a compensation signal is directly generated through digital synthesis technology. Then, the modulated signal is combined with the compensation signal and transmitted to the B channel of the low noise amplifier. After multiple iterations, when the baseband signal strength is finally equal to the background noise level, it indicates that the compensation effect has achieved the expected goal, and the iteration process ends.

[0015] Furthermore, in S2, an excitation signal is generated by an excitation field, the excitation signal is coupled with the receiving coil to generate a feedthrough signal, and the receiving coil adopts a first-order gradiometer design to passively compensate for the feedthrough signal. This sentence is changed as follows; both the transmitting coil and the receiving coil are wound with Pack Litz wire. Preferably, in one embodiment, the transmitting coil is designed as a single-layer structure, and is wound with 0.5mm×600 strands of Pack Litz wire; the receiving coil is in the form of a one-dimensional gradient coil, and the receiving coil adopts a two-section layout, with a compensation coil on the left and a detection coil on the right, and their winding directions are set to be opposite to achieve a specific electromagnetic effect.

[0016] Furthermore, the excitation signal collected in S3 can be expressed as:

[0017] U(ωt,φ)=A E cos(ωt,φ E )

[0018] Among them A E cos(ωt,φ E) is the feedthrough signal after passive compensation. Since the frequency of the feedthrough signal is consistent with the excitation signal frequency generated by the FPGA, the amplitude A of the feedthrough signal is calculated by the phase sensitive detection (PSD) algorithm. E With phase φ E , and use them as the amplitude and initial phase of the compensation signal.

[0019] Furthermore, the phase sensitive detection (PSD) algorithm uses an orthogonal lock-in amplifier to calculate the amplitude and phase of the feedthrough, where S I (t) = A I sin(ωt+φ)+B(t) is the input signal, A I sin(ωt+φ) is the feedthrough signal, B(t) is the total noise, and the reference signal is divided into two parts with a phase difference of 90°, which is:

[0020]

[0021] After PSD, we can get:

[0022]

[0023] Furthermore, the S pd1 The output is divided into three parts. The first part is a DC signal. The second part is a signal with a frequency twice that of the reference signal (i.e., 2 times the frequency), which is filtered out by a low-pass filter (LPF). The third part is the product of noise and the reference signal. Because the sinusoidal signal is periodic and has no correlation with the noise signal, this integral is 0.

[0024] Furthermore, after passing through the low-pass filter (LPF), we get:

[0025]

[0026] Take the square root of the sum of the squares of X and Y and find the inverse tangent:

[0027]

[0028] Where A1 is the amplitude of the feedthrough signal, and θ is the phase of the feedthrough signal, from which the amplitude and phase of the feedthrough signal can be calculated.

[0029] (III) Beneficial effects

[0030] Compared with the prior art, the present invention provides a method for compensating the fundamental frequency signal of magnetic nanoparticle imaging based on FPGA, which has the following beneficial effects:

[0031] The present invention uses FPGA real-time calculation and anti-phase signal modulation to accurately offset the feedthrough signal. The efficient computing capability of FPGA supports real-time signal processing and adapts to changes in the excitation magnetic field. Through an iterative compensation process, the fundamental frequency component of the particle signal is restored to ensure signal integrity. The digital method reduces hardware complexity and improves system stability and repeatability. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a schematic diagram of the system flow of the present invention;

[0033] Figure 2 It is a schematic diagram of the algorithm flow of the present invention;

[0034] Figure 3 It is the basic block diagram of the phase sensitive detection (PSD) of the present invention;

[0035] Figure 4 This is a comparison diagram of the time domain waveforms of the feedthrough signal before and after active compensation of the present invention;

[0036] Figure 5 It is a schematic diagram of the transmitting coil of the present invention;

[0037] Figure 6 Schematic diagram of the receiving coil of the present invention. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] Example

[0040] like Figure 1-6 As shown, an embodiment of the present invention provides a method for compensating a fundamental frequency signal of magnetic nanoparticle imaging based on FPGA, comprising the following steps:

[0041] S1: The field programmable gate array (FPGA) sends an excitation signal to the power amplifier; the excitation signal frequency is between 1kHz and 100kHz, and the model of the FPGA development board is: (ALINX Black Gold AX7020);

[0042] S2: The power amplifier transmits the signal to the transmitting coil, the transmitting coil generates an excitation field, the first-order gradiometer receives the signal, and passively compensates the received feedthrough signal, wherein the power amplifier model is: (AE Techron, 7224, USA);

[0043] S3: The passively compensated feedthrough signal is first converted into a digital signal through an analog-to-digital converter (ADC), and then the digital signal is transmitted to the FPGA. The amplitude and phase of the feedthrough signal are further calculated using a dynamic feedthrough compensation algorithm. The ADC module model is: (AN9238);

[0044] S4: The FPGA generates a modulation signal with a phase opposite to that of the feedthrough signal and transmits the signal to a digital-to-analog converter (DAC). The DAC then sends the converted analog modulation signal to a differential low-noise amplifier.

[0045] S5: Repeat steps S3 and S4 until the strength of the acquired baseband signal is equivalent to the noise floor level, which indicates that the compensation effect has achieved the expected goal;

[0046] S6: Adjust the gain of the low noise amplifier. The feedthrough signal is first amplified by the A channel of the low noise amplifier and then subtracted from the modulation signal. The difference signal is sent to the phase-locked amplifier to calculate the strength of the baseband signal. If the calculated baseband signal strength is not equivalent to the background noise level, then the calculation result is divided by the gain, and the compensation signal is directly generated through digital synthesis technology. Then, the modulation signal and the compensation signal are combined and transmitted to the B channel of the low noise amplifier. After multiple iterations, when the baseband signal strength is finally equal to the background noise level, it indicates that the compensation effect has achieved the expected goal, and the iteration process ends. The low noise amplifier model is: (LNA, Stanford Research Systems, SR560, USA), and the DAC module model is: (AN9767).

[0047] In S2, the excitation field generates an excitation signal, which is coupled with the receiving coil to generate a feedthrough signal. The first-order gradiometer passively compensates the feedthrough signal. Both the transmitting coil and the receiving coil are wound with high-quality PackLitz wire. This choice is intended to effectively avoid the adverse effects of the skin effect on the signal quality. The specific structure is as follows: Figure 5 As shown in the figure, this single-layer design strategy is based on the study of the potential defects of the multi-layer structure. The multi-layer structure often leads to a decrease in the resonant frequency of the coil, which may not only aggravate the interference of floor noise, but also significantly weaken the signal-to-noise ratio of the system, thus being detrimental to the realization of high-precision imaging. The receiving coil adopts the form of a one-dimensional gradient coil, and its specific structure is as follows Figure 6As shown, the receiving coil adopts a two-section layout, with the compensation coil on the left and the detection coil on the right, and their winding directions are set to be opposite to achieve a specific electromagnetic effect. The core concept of this design is that the compensation coil can induce a voltage in the receiving coil that offsets the voltage directly induced by the excitation field in the detection coil, thereby realizing an efficient passive compensation mechanism. In theory, this mechanism can greatly reduce the interference of the feedthrough signal on the detection signal, thereby improving the measurement accuracy of the system. However, it is worth noting that in the actual winding process, due to factors such as the unevenness of the winding, it may not be possible to achieve a perfect compensation effect. Therefore, a fine-tuning strategy is proposed to further optimize the compensation effect by fine-tuning the coil parameters to ensure that the system can show the best performance in practical applications; through this coil group design, the inductive decoupling of the transmitting coil and the receiving coil is achieved, which significantly reduces the influence of the feedthrough signal caused by the excitation field, laying the foundation for subsequent signal processing;

[0048] The signal collected by the dynamic compensation feedthrough algorithm in S3 can be expressed as:

[0049] U(ωt,φ)=A E cos(ωt,φ E )

[0050] Among them A E cos(ωt,φ E ) is the feedthrough signal after passive compensation. Since the frequency of the feedthrough signal is consistent with the excitation signal frequency generated by the FPGA, the amplitude A of the feedthrough signal needs to be calculated using the phase-sensitive detection (PSD) algorithm. E With phase φ E , and use them as the amplitude and initial phase of the compensation signal. Phase-sensitive detection is a technology that uses the phase information of the signal to improve the accuracy of signal detection. It enhances the signal response and effectively suppresses noise by comparing the phase of the input signal with the phase of the reference signal. Phase-sensitive detection can be regarded as a bandpass filter with a very narrow bandwidth. The basic block diagram of phase-sensitive detection (PSD) is shown in Figure 3 shown.

[0051] S I (t) is the output signal + noise, S R (t) is the reference signal. They are signals with the same frequency. The phase-sensitive detection (PSD) algorithm uses an orthogonal lock-in amplifier to calculate the amplitude and phase of the feedthrough. I (t) = A I sin(ωt+φ)+B(t) is the input signal, A Isin(ωt+φ) is the feedthrough signal, B(t) is the total noise, and the reference signal is divided into two parts with a phase difference of 90°, which is:

[0052]

[0053] After PSD, we can get:

[0054]

[0055] S pd0 The output is divided into three parts, the first part is a DC signal, which represents the constant component in the output.

[0056] The second part is a signal with a frequency twice that of the reference signal (i.e., 2 times the frequency). This 2 times frequency signal can be effectively filtered out by a low pass filter (LPF), because the low pass filter allows low frequency signals to pass and blocks high frequency signals, so high frequency components such as 2 times the frequency will be attenuated to a negligible level.

[0057] The third part is composed of the product of the noise signal and the reference signal. Since the reference signal is a sine wave with clear periodicity, while the noise signal is random and non-periodic, there is no correlation between the two. When performing correlation operations or integration processing, the integral result of the product of such uncorrelated signals tends to zero.

[0058] After passing through a low-pass filter (LPF), we get:

[0059]

[0060] Take the square root of the sum of the squares of X and Y and find the inverse tangent:

[0061]

[0062] Where A1 is the amplitude of the feedthrough signal, and θ is the phase of the feedthrough signal, from which the amplitude and phase of the feedthrough signal can be calculated.

[0063] Figure 4 The compensation performance comparison between the prior art and the method proposed in the present invention is shown.

[0064] Compared with other existing methods, the method proposed in this invention significantly improves the attenuation effect and stability and simplifies the compensation process. The amplitude of the feedthrough signal before active compensation is -16.61dBV, and the amplitude of the feedthrough signal after active compensation is -58.71dBV. The attenuation degree of the feedthrough signal is 42.1dBV. This result effectively verifies that the active compensation method proposed in this paper can achieve significant attenuation of the feedthrough signal. More accurate compensation of the feedthrough signal improves compensation efficiency and accuracy.

[0065] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for compensating fundamental frequency signals of magnetic nanoparticle imaging based on FPGA, characterized in that: The steps include: S1: The field programmable gate array (FPGA) sends an excitation signal to the power amplifier; the excitation signal frequency is between 1kHz and 100kHz; S2: The power amplifier then transmits the signal to the transmitting coil, which generates an excitation field. The first-order gradiometer receives the signal and passively compensates the received feedthrough signal. S3: The passively compensated feedthrough signal is first converted into a digital signal through an analog-to-digital converter (ADC), and then the digital signal is transmitted to the FPGA, and the amplitude and phase of the feedthrough signal are further calculated using a dynamic feedthrough compensation algorithm; S4: The FPGA generates a modulation signal with a phase opposite to that of the feedthrough signal and transmits the signal to a digital-to-analog converter (DAC). The DAC then sends the converted analog modulation signal to a differential low-noise amplifier. S5: Repeat steps S3 and S4 until the strength of the acquired baseband signal is equivalent to the noise floor level, which indicates that the compensation effect has achieved the expected goal; S6: Adjust the gain of the low noise amplifier. The feedthrough signal is first amplified by the A channel of the low noise amplifier and then subtracted from the modulated signal. The difference signal is sent to the phase-locked amplifier to calculate the strength of the baseband signal. If the calculated baseband signal strength is not equivalent to the background noise level, then the calculation result is divided by the gain, and a compensation signal is directly generated through digital synthesis technology. Then, the modulated signal is combined with the compensation signal and transmitted to the B channel of the low noise amplifier. After multiple iterations, when the baseband signal strength is finally equal to the background noise level, it indicates that the compensation effect has achieved the expected goal, and the iteration process ends.

2. The method for compensating the fundamental frequency signal of magnetic nanoparticle imaging based on FPGA according to claim 1, characterized in that: In S2, an excitation signal is generated by an excitation field, and the excitation signal is coupled with a receiving coil to generate a feedthrough signal. The receiving coil adopts a first-order gradiometer design to passively compensate the feedthrough signal. This sentence is changed as follows; both the transmitting coil and the receiving coil are wound with Pack Litz wire. Preferably, in one embodiment, the transmitting coil is designed as a single-layer structure, and is wound with 0.5 mm×600 strands of Pack Litz wire; The receiving coil is in the form of a one-dimensional gradient coil, and the receiving coil adopts a two-section layout, with a compensation coil on the left and a detection coil on the right, and their winding directions are set to be opposite to achieve a specific electromagnetic effect.

3. The method for compensating fundamental frequency signals of magnetic nanoparticle imaging based on FPGA according to claim 1, characterized in that: The excitation signal collected in S3 can be expressed as: U(ωt,φ)=A E cos(ωt,φ E ) Among them A E cos(ωt,φ E ) is the feedthrough signal after passive compensation. Since the frequency of the feedthrough signal is consistent with the excitation signal frequency generated by the FPGA, the amplitude A of the feedthrough signal is calculated by the phase sensitive detection (PSD) algorithm. E With phase φ E , and use them as the amplitude and initial phase of the compensation signal.

4. The method for compensating the fundamental frequency signal of magnetic nanoparticle imaging based on FPGA according to claim 3, characterized in that: The phase sensitive detection (PSD) algorithm uses an orthogonal lock-in amplifier to calculate the amplitude and phase of the feedthrough, where S I (t) = A I sin(ωt+φ)+B(t) is the input signal, A I sin(ωt+φ) is the feedthrough signal, B(t) is the total noise, and the reference signal is divided into two parts with a phase difference of 90°, which is: After PSD, we can get:

5. The method for compensating fundamental frequency signals of magnetic nanoparticle imaging based on FPGA according to claim 4, characterized in that: The S pd1 The output is divided into three parts. The first part is a DC signal. The second part is a signal with a frequency twice that of the reference signal (i.e., 2 times the frequency), which is filtered out by a low-pass filter (LPF). The third part is the product of noise and the reference signal. Because the sinusoidal signal is periodic and has no correlation with the noise signal, this integral is 0.

6. The method for compensating fundamental frequency signals of magnetic nanoparticle imaging based on FPGA according to claim 5, characterized in that: After passing through the low-pass filter (LPF), we get: Take the square root of the sum of the squares of X and Y and find the inverse tangent: Where A1 is the amplitude of the feedthrough signal, and θ is the phase of the feedthrough signal, from which the amplitude and phase of the feedthrough signal can be calculated.

Images