HYSTERESIS EFFECT-BASED FIELD FREE POINT-MAGNETIC PARTICLE IMAGING METHOD

Abstract

A hysteresis effect-based Field Free Point-Magnetic Particle Imaging (FFP-MPI) method includes the following steps: acquiring a hysteresis loop model of Superparamagnetic Iron Oxide Nanoparticles (SPIOs); calculating to obtain a Point Spread Function (PSF) of the SPIOs on the basis of a sinusoidal excitation magnetic field and the hysteresis loop model of the SPIOs; acquiring an original reconstructed image of FFP-MPI on the basis an FFP moving track and a voltage signal; performing deconvolution on the original image with respect to the PSF considering an hysteresis effect, so as to obtain a final reconstructed image; the artifacts and phase errors of image reconstruction caused by the hysteresis effect of the SPIOs with large particle sizes are reduced, the deficiency in reconstruction by the traditional reconstruction method that ignores the hysteresis effect is overcome, the reconstruction speed and the resolution are greatly improved, and the application range of the SPIOs is expanded.

Claims

1. A hysteresis effect-based Field Free Point-Magnetic Particle Imaging (FFP-MPI) method, comprising the following steps: S1: acquiring a hysteresis loop model of Superparamagnetic Iron Oxide Nanoparticles (SPIOs); S2: calculating to obtain a Point Spread Function (PSF) of the SPIOs on basis of a sinusoidal excitation magnetic field and the hysteresis loop model of the SPIOs; S3: acquiring an original reconstructed image of FFP-MPI on the basis of an FFP moving track and a voltage signal; S4: performing deconvolution on the original reconstructed image of FFP-MPI with respect to the PSF considering hysteresis effect to obtain a final reconstructed image.

2. The method according to claim 1, the step for acquiring the hysteresis loop of the SPIOs in S1 further comprising: measuring a plurality of sets of feature point data of the SPIOs in an Alternating Current (AC) magnetic field, substituting the plurality of sets of feature point data into an M-H hysteresis curve model, solving to obtain parameters: saturation magnetization vector M.sub.s, magnetic field coupling strength α, magnetic domain density a, average energy k, and magnetization reversibility c, and substituting said parameters into the M-H hysteresis curve model to obtain the hysteresis loop of the SPIOs.

3. The method according to claim 2, wherein the M-H hysteresis curve model is: $\{\begin{array}{c}\frac{\mathrm{dM}}{\mathrm{dH}}=\frac{1}{\left(1+c\right)}\frac{\left(M_1-M\right)}{\delta k/{\mu }_0-\alpha \left(M_1-M\right)}+\frac{c}{\left(1+c\right)}\frac{{\mathrm{dM}}_1}{\mathrm{dH}}\\ M_1=M_s(\coth \left(\frac{H+\mathrm{\alpha M}}{a}\right)-\frac{a}{H+\mathrm{\alpha M}})\end{array}$ wherein, H is an externally applied excitation magnetic field, M is a magnetization vector of the SPIOs, and μ.sub.0 is the permeability of vacuum; wherein when the externally applied excitation magnetic field increases positively, δ=1; wherein when the externally applied excitation magnetic field decreases positively, δ=−1.

4. The method according to claim 1, the step for calculating to obtain the PSF of the SPIOs in S2 further comprising: Wherein when the SPIOs are excited by a sinusoidal excitation magnetic field,
H(t)=A cos(ωt) wherein, t is time, A is a magnetic field amplitude value, and ω is an angular frequency of an excitation magnetic field; wherein the above formula is substituted into the hysteresis loop model of the SPIOs to obtain a function M(t) of a magnetization vector of the SPIOs along with the time; wherein a derivative of M(t) is taken with respect to the time, and the obtained PSF of the SPIOs is as follows: $\mathrm{PSF}=\frac{\mathrm{dM}\left(t\right)}{\mathrm{dt}}=A\mathrm{\omega sin}\left(\mathrm{\omega t}\right)(\frac{1}{\left(1+c\right)}\frac{\left(M_1-M\right)}{\delta k/{\mu }_0-\alpha \left(M_1-M\right)}+\frac{c}{\left(1+c\right)}\frac{{\mathrm{dM}}_1}{\mathrm{dH}}).$

5. The method according to claim 1, the step for acquiring the original reconstructed image of the FFP-MPI on the basis of the FFP moving track and the voltage signal in S3 further comprising: moving an FFP according to a scanning track, wherein the moving speed is v, and the position is r; scanning the overall view field by using the MPI to obtain a voltage signal u(t) of an induction coil, wherein a relationship between the original reconstructed image and the voltage signal is as follows: ${\mathrm{IMG}}_{\mathrm{raw}}=u\left(t\right)/v=c\left(r\right)***\mathrm{PSF}$ wherein, c(r) is a variation distribution matrix of the concentration along with the position, and *** is a three-dimensional convolution symbol; dividing the voltage signal by a scanning speed, and splicing the image according to the scanning track to obtain the original reconstructed image.

6. The method according to claim 1, the step for performing deconvolution on the original reconstructed image with respect to the PSF considering the hysteresis effect to obtain a final reconstructed image in S4 is: $\mathrm{IMG}\text{?}=\mathrm{IMG}\text{?}\mathrm{PSF};$ $\text{?}\text{indicates text missing or illegible when filed}$ wherein {tilde over (*)}{tilde over (*)}{tilde over (*)} is three-dimensional deconvolution.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0029] FIG. 1 illustrates a flowchart of a hysteresis effect-based FFP-MPI method.

[0030] FIG. 2 illustrates an M-H hysteresis loop considering a hysteresis effect and a magnetization curve ignoring the hysteresis effect in an ideal situation.

[0031] FIG. 3 illustrates magnetization vector variation curves along with time considering the hysteresis effect and ignoring the hysteresis effect in the ideal situation.

[0032] FIG. 4 illustrates PSF variation curves along with time considering the hysteresis effect and ignoring the hysteresis effect in the ideal situation.

[0033] FIG. 5 illustrates an original reconstructed image of FFP-MPI.

[0034] FIG. 6 illustrates a final reconstructed image of FFP-MPI.

DETAILED DESCRIPTION OF THE EMBODIMENTS

[0035] The present disclosure will be further described in detail below with reference to accompanying drawings and embodiments.

[0036] A hysteresis effect-based FFP-MPI method includes the following steps.

[0037] S1: a hysteresis loop of SPIOs is acquired by measuring the parameters through an MPS and combining an M-H curve model of a hysteresis effect.

[0038] The SPIOs are excited by a high-frequency (20-45 kHz) sinusoidal excitation magnetic field in an MPI device, and a magnetization process of the SPIOs follows an M-H hysteresis curve model:

[00005]$\begin{array}{cc}\{\begin{array}{c}\frac{\mathrm{dM}}{\mathrm{dH}}=\frac{1}{\left(1+c\right)}\frac{\left(M_1-M\right)}{\delta k/{\mu }_0-\alpha \left(M_1-M\right)}+\frac{c}{\left(1+c\right)}\frac{{\mathrm{dM}}_1}{\mathrm{dH}}\\ M_1=M_s(\coth \left(\frac{H+\mathrm{\alpha M}}{a}\right)-\frac{a}{H+\mathrm{\alpha M}})\end{array}& \left(1.1\right)\end{array}$

[0039] Where, H is an externally applied excitation magnetic field, M is a magnetization vector of the SPIOs, and μ.sub.0 is the permeability of vacuum. In the model of Formula (1.1), when the externally applied excitation magnetic field increases positively, δ=1; when the externally applied excitation magnetic field decreases positively, δ=−1. Three sets of data of the SPIOs in an AC magnetic field can be measured through the MPS: the residual hysteresis M.sub.r when H=0, the coercive field intensity H.sub.c when M=0, and the maximum magnetization vector M.sub.max at the maximum magnetic field intensity H.sub.max. The three sets of data are substituted into (1.1) to solve the following parameters: saturation magnetization vector M.sub.s, magnetic field coupling strength α, magnetic domain density a, average energy k, and magnetization reversibility c. The parameters are substituted into the M-H hysteresis curve model to obtain the hysteresis loop of the SPIOs. For example, when the diameter of the SPIOs is 30 nm, the hysteresis loop considering the hysteresis effect is shown as the hysteresis loop with the hysteresis effect in FIG. 2, and the magnetization curve not considering the hysteresis effect in an ideal situation is shown as the hysteresis curve without the hysteresis effect in FIG. 2.

[0040] S2: a response magnetization vector of the SPIOs is obtained in combination with the hysteresis loop of the SPIOs on the basis of the sinusoidal excitation magnetic field, and a derivative is solved with respect to time, so as to obtain a PSF of the SPIOs.

[0041] When the SPIOs are excited by the sinusoidal excitation magnetic field:

H(t)=A cos(ωt)  (1.2)

[0042] Where, t is time, A is a magnetic field amplitude value, and w is an angular frequency of an excitation magnetic field. Formula (1.2) is substituted into Formula (1.1) to obtain a function M(t) of the magnetization vector of the SPIOs along with time. As shown in FIG. 3, the variation curve of the magnetization vector considering the hysteresis effect along with time is as shown in a curve considering the hysteresis effect, and the variation of the magnetization vector ignoring the hysteresis effect in an ideal situation is shown as the curve without the hysteresis effect.

[0043] PSF is a derivative of the magnetization vector with respect to time, so:

[00006]$\begin{array}{cc}\mathrm{PSF}=\frac{\mathrm{dM}\left(t\right)}{\mathrm{dt}}=A\mathrm{\omega sin}\left(\mathrm{\omega t}\right)(\frac{1}{\left(1+c\right)}\frac{\left(M_1-M\right)}{\delta k/{\mu }_0-\alpha \left(M_1-M\right)}+\frac{c}{\left(1+c\right)}\frac{{\mathrm{dM}}_1}{\mathrm{dH}})& \left(1.3\right)\end{array}$

[0044] The variation curve of PSF along with time is as shown in FIG. 4. The curve with the hysteresis effect represents the PSF considering the hysteresis effect. Compared with a hysteresis effect-free PSF ignoring the hysteresis effect, the phase moves backward, and the shapes of the two curves are the same.

[0045] S3: an original reconstructed image of the FFP-MPI is acquired on the basis of an FFP moving track and a voltage signal.

[0046] The scanning track of the FFP generally includes: a Cartesian scanning track and a Lissajous scanning track. The FFP is moved according to a certain scanning track, where the moving speed is v, and the position varying along with time is r(t); the view field of the MPI is scanned once to obtain a voltage signal u(t) of an induction coil. A relationship between the original reconstructed image and the voltage signal is as follows:

IMG.sub.raw=u(t)/v=c(r)***PSF  (1.4)

[0047] Where, c(r) is a variation distribution matrix of concentration along with the position, and *** is a three-dimensional convolution symbol. The voltage signal is divided by a scanning speed, and the image is spliced according to the scanning track to obtain the original reconstructed image. At present, an image result used in the MPI field is the original reconstructed image. The original reconstructed image can reflect the general situation of concentration distribution of the SPIOs, but not an accurate concentration distribution result. In addition, the resolution is poor, the phase backward shift of the PSF caused by the hysteresis effect is ignored, and further image optimization is needed. For example, MPI scans two parallel strip samples, and the original reconstructed image is as shown in FIG. 5.

[0048] S4: deconvolution is performed on the original reconstructed image through the PSF considering the hysteresis effect, so as to obtain a final reconstructed image.

[0049] The original reconstructed image ignores the phase backward shift of the hysteresis effect of the SPIOs, and a position error is generated during corresponding from a time domain voltage signal to each FFP. Deconvolution is performed on the original reconstructed image with respect to the PSF, a phase backward shift error caused by the hysteresis effect is corrected by using the PSF, and the result after the deconvolution is a final reconstructed result.

[00007]$\begin{array}{cc}\mathrm{IMG}\text{?}=\mathrm{IMG}\text{?}\mathrm{PSF}& \left(1.5\right)\end{array}$$\text{?}\text{indicates text missing or illegible when filed}$

[0050] Where, {tilde over (*)}{tilde over (*)}{tilde over (*)} is three-dimensional deconvolution. For example, MPI scans two parallel strip samples, and the final reconstructed image after the deconvolution is as shown in FIG. 6.

[0051] The above is merely a specific implementation mode of the present disclosure and is not intended to limit the scope of protection of the present disclosure. Any modifications, equivalent replacements, improvements and the like made within the spirit and principle of the present disclosure shall fall within the scope of protection of the present disclosure.