IEEE Trans Biomed Eng. 2023 Jul;70(7):2035-2045. doi: 10.1109/TBME.2023.3234256. Epub 2023 Jun 19.
Magnetic particle imaging (MPI) is a promising medical modality that can image superparamagnetic iron-oxide nanoparticle (SPIO) concentration distributions safely and with high sensitivity. In the x-space reconstruction algorithm, the Langevin function is inaccurate in modeling the dynamic magnetization of SPIOs. This problem prevents the x-space algorithm from achieving a high spatial resolution reconstruction.
We propose a more accurate model to describe the dynamic magnetization of SPIOs, named the modified Jiles-Atherton (MJA) model, and apply it to the x-space algorithm to improve the image resolution. Considering the relaxation effect of SPIOs, the MJA model generates the magnetization curve via an ordinary differential equation. Three modifications are also introduced to further improve its accuracy and robustness.
In magnetic particle spectrometry experiments, the MJA model shows higher accuracy than the Langevin and Debye models under various test conditions. The average root-mean-square error is 0.055, 83% and 58% lower than the Langevin and Debye models, respectively. In MPI reconstruction experiments, the MJA x-space improves the spatial resolution by 64% and 48% compared to the x-space and Debye x-space methods, respectively.
The MJA model shows high accuracy and robustness in modeling the dynamic magnetization behavior of SPIOs. By integrating the MJA model into the x-space algorithm, the spatial resolution of MPI technology was improved.
Improving the spatial resolution by using the MJA model helps MPI have a better performance in medical fields, including cardiovascular imaging.
磁共振粒子成像（MPI）是一种很有前途的医学成像方式，可以安全且高灵敏度地对超顺磁氧化铁纳米颗粒（SPIO）的浓度分布进行成像。在 x 空间重建算法中，朗之万函数在对 SPIO 的动态磁化进行建模时不够准确。这个问题导致 x 空间算法无法实现高空间分辨率的重建。
我们提出了一种更准确的模型来描述 SPIO 的动态磁化，称为修正的 Jiles-Atherton（MJA）模型，并将其应用于 x 空间算法以提高图像分辨率。考虑到 SPIO 的弛豫效应，MJA 模型通过常微分方程生成磁化曲线。此外，还引入了三个改进措施来进一步提高其准确性和鲁棒性。
在磁粒子谱实验中，在各种测试条件下，MJA 模型比朗之万和德拜模型具有更高的准确性。平均均方根误差分别比朗之万和德拜模型低 0.055、83%和 58%。在 MPI 重建实验中，与 x 空间和 Debye x 空间方法相比，MJA x 空间分别将空间分辨率提高了 64%和 48%。
MJA 模型在建模 SPIO 动态磁化行为方面具有高精度和高鲁棒性。通过将 MJA 模型集成到 x 空间算法中，提高了 MPI 技术的空间分辨率。
通过使用 MJA 模型提高空间分辨率有助于 MPI 在心血管成像等医学领域获得更好的性能。
