Spectral Spatial RF Pulses for Gradient Echo fMRI

描述（由申请人提供）：“梯度回波MRI的频谱空间射频脉冲”提案是一个MRI技术开发项目，旨在设计使用单台和多台3T发射器的频谱空间射频（RF）激励。这些脉冲的设计目标是抑制不需要的脂质信号，减少敏感性伪影，并改善梯度回声MRI的切片剖面（B1+）均匀性。梯度回波应用，如血氧水平依赖（BOLD）脑功能MRI (fMRI)，由于磁化率的变化，下脑区域存在较大的信号空洞。此外，良好的梯度回波对比度所需的高场使图像容易因B1+不均匀性而发生强度变化。解决这些限制的方法对于充分利用MRI的优势来改善医疗保健和研究是很重要的。我们首先在一个或多个发射机上提出了用于切片和频率选择性的二维频谱空间脉冲。这些脉冲可用于抑制脂质和消除通过平面的敏感性梯度。磁化率伪影校正假设通面梯度是非共振频率的函数。获取场图以确定通过平面梯度和非共振的空间分布将验证这一假设。然后利用平行传输方法利用地图的空间变化。下一种方法将是为平行发射器设计4D光谱空间脉冲，以开发同时校正平面内磁化伪影、平面内发射器（B1+）不均匀性并提供脂质抑制的激励。然后，脉冲生成算法将被移植到图形编程单元（gpu）上，以提高速度。这些脉冲将通过模拟、幻影和人类控制梯度回波成像研究进行测试和表征。脉冲的最终验证将使用人类对照扫描与敏感性加权成像（SWI）， T2*映射和屏气BOLD功能磁共振成像实验。本提案中描述的方法的成功开发将克服梯度回声MRI的主要限制，使以前不可能实现的广泛临床应用成为可能。此外，频谱空间脉冲和并行发射机的应用是本方案的新颖之处，是多维射频脉冲设计的一大进步。

DESCRIPTION (provided by applicant): This proposal "Spectral-Spatial RF Pulses for Gradient Echo MRI" is an MRI technology development project to design spectral-spatial Radio Frequency (RF) excitations using single and multiple transmitters at 3T. These pulses will be designed with the goal of suppressing unwanted lipid signal, reducing susceptibility artifacts, and improving slice profile (B1+) uniformity in gradient echo MRI. Gradient echo applications such as blood oxygen level dependent (BOLD) brain functional MRI (fMRI) are plagued by large signal voids in the inferior brain regions due to magnetic susceptibility variations. Furthermore, the high fields required for good gradient echo contrast make the images prone to intensity variations from B1+ inhomogeneity. Methods that address these limitations are important to exploit the full benefits of MRI for improved health care and research. We first propose 2D spectral spatial pulses for slice and frequency selectivity on one or multiple transmitters. These pulses can be used for both lipid suppression and the cancellation of the through-plane susceptibility gradient. The susceptibility artifact correction assumes that the through-plane gradient is a function of off- resonance frequency. Acquiring field maps to determine the spatial distribution of through-plane gradients and off-resonance will test this assumption. The spatial variations of the maps will then be exploited using parallel transmission methods. The next approach will be to design 4D spectral-spatial pulses for parallel transmitters to develop excitations that simultaneously correct for through-plane susceptibility artifact, in-plane transmitter (B1+) inhomogeneity, and provide lipid suppression. The pulse generation algorithms will then be ported for use on graphics programming units (GPUs) for increased speed. The pulses will be tested and characterized with simulations and phantom and human control gradient echo imaging studies. Final validation of the pulses will use human control scanning with susceptibility weighted imaging (SWI), T2* mapping, and breath-holding BOLD fMRI experiments. Success in developing the methods described in this proposal will overcome major limitations in gradient echo MRI, making feasible a broad range of clinical applications not previously possible. Furthermore, the application spectral-spatial pulses and parallel transmitters is novel to this proposal and represent a big step forward in multi-dimensional RF pulse design. PUBLIC HEALTH RELEVANCE: Magnetic resonance imaging (MRI) is a powerful and non-invasive technique for observing anatomy, structure, and function in the human body. In particular gradient echo MRI is useful for a number of applications including brain functional and structural imaging. However, the high fields required for adequate gradient echo contrast also produce challenges and obstacles in the form of image artifacts. The goal of this project is to develop and validate a system of techniques to correct for these field related MRI artifacts. The proposed research will ultimately aid in reducing the cost and duration of MRI examinations and provide improved diagnostic accuracy.