Department of Imaging, Lurie Family Imaging Center, Center for Biomedical Imaging in Oncology, Dana-Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts, United States of America.
Neuroimaging Research, Radiology Department, Boston Children's Hospital, Harvard Medical School, Boston, Massachusetts, United States of America.
PLoS One. 2018 Jul 23;13(7):e0200611. doi: 10.1371/journal.pone.0200611. eCollection 2018.
Magnetic Resonance Imaging (MRI) relies on optimal scanning parameters to achieve maximal signal-to-noise ratio (SNR) and high contrast-to-noise ratio (CNR) between tissues resulting in high quality images. The optimization of such parameters is often laborious, time consuming, and user-dependent, making harmonization of imaging parameters a difficult task. In this report, we aim to develop and validate a computer simulation technique that can reliably provide "optimal in vivo scanning parameters" ready to be used for in vivo evaluation of disease models.
A glioblastoma murine model was investigated using several MRI imaging methods. Such MRI methods underwent a simulated and an in vivo scanning parameter optimization in pre- and post-contrast conditions that involved the investigation of tumor, brain parenchyma and cerebrospinal fluid (CSF) CNR values in addition to the time relaxation values of the related tissues. The CNR tissues information were analyzed and the derived scanning parameters compared in order to validate the simulated methodology as a reliable technique for "optimal in vivo scanning parameters" estimation.
The CNRs and the related scanning parameters were better correlated when spin-echo-based sequences were used rather than the gradient-echo-based sequences due to augmented inhomogeneity artifacts affecting the latter methods. "Optimal in vivo scanning parameters" were generated successfully by the simulations after initial scanning parameter adjustments that conformed to some of the parameters derived from the in vivo experiment.
Scanning parameter optimization using the computer simulation was shown to be a valid surrogate to the in vivo approach in a glioblastoma murine model yielding in a better delineation and differentiation of the tumor from the contralateral hemisphere. In addition to drastically reducing the time invested in choosing optimal scanning parameters when compared to an in vivo approach, this simulation program could also be used to harmonize MRI acquisition parameters across scanners from different vendors.
磁共振成像（MRI）依赖于最佳扫描参数，以实现最大信号噪声比（SNR）和组织间高的对比噪声比（CNR），从而获得高质量的图像。这些参数的优化通常是费力、耗时且依赖于用户的，因此实现成像参数的协调是一项艰巨的任务。在本报告中，我们旨在开发和验证一种计算机模拟技术，该技术能够可靠地提供“最佳的活体扫描参数”，以便用于活体评估疾病模型。
使用几种 MRI 成像方法对胶质母细胞瘤鼠模型进行了研究。这些 MRI 方法在对比前后经历了模拟和活体扫描参数优化，除了相关组织的时间弛豫值外，还涉及肿瘤、脑实质和脑脊液（CSF）CNR 值的研究。分析了 CNR 组织信息，并对衍生的扫描参数进行了比较，以验证模拟方法作为“最佳活体扫描参数”估计的可靠技术。
由于梯度回波序列的不均匀性伪影影响了后者方法，因此基于自旋回波的序列比基于梯度回波的序列具有更好的 CNR 和相关扫描参数相关性。通过模拟成功生成了“最佳活体扫描参数”，这些参数是在初始扫描参数调整后生成的，这些参数符合一些从活体实验中得出的参数。
在胶质母细胞瘤鼠模型中，使用计算机模拟进行扫描参数优化被证明是活体方法的有效替代方法，可更好地从对侧半球区分和分化肿瘤。与活体方法相比，这种模拟程序除了大大减少了选择最佳扫描参数所需的时间外，还可以用于协调来自不同供应商的不同扫描仪的 MRI 采集参数。
