Sir Peter Mansfield Magnetic Resonance Centre, School of Physics and Astronomy, University of Nottingham, Nottingham, UK.
Sir Peter Mansfield Magnetic Resonance Centre, School of Physics and Astronomy, University of Nottingham, Nottingham, UK.
Neuroimage. 2014 Feb 1;86:35-42. doi: 10.1016/j.neuroimage.2013.04.077. Epub 2013 Apr 29.
State of the art magnetic resonance imaging (MRI) scanners are generally equipped with multi-element receive coils; 16 or 32 channel coils are common. Their development has been predominant for parallel imaging to enable faster scanning. Less consideration has been given to localized magnetic resonance spectroscopy (MRS). Multinuclear studies, for example (31)P or (13)C MRS, are often conducted with a single element coil located over the region of interest. (1)H MRS studies have generally employed the same multi-element coils used for MRI, but little consideration has been given as to how the spectroscopic data from the different channels are combined. In many cases it is simply co-added with detrimental effect on the signal to noise ratio. In this study, we derive the optimum method for combining multi-coil data, namely weighting with the ratio of signal to the square of the noise. We show that provided that the noise is uncorrelated, this is the theoretical optimal combination. The method is demonstrated for in vivo proton MRS data acquired using a 32 channel receive coil at 7T in four different brain areas; left motor and right motor, occipital cortex and medial frontal cortex.
最先进的磁共振成像（MRI）扫描仪通常配备有多元素接收线圈；16 或 32 通道线圈很常见。它们的发展主要是为了实现并行成像，从而实现更快的扫描。而对于局部磁共振波谱（MRS）的考虑则较少。例如，多核研究（例如 31P 或 13C MRS）通常使用位于感兴趣区域上方的单个元素线圈进行。（1）H MRS 研究通常使用用于 MRI 的相同多元素线圈，但很少考虑如何组合来自不同通道的光谱数据。在许多情况下，它只是简单地相加，这对信噪比有不利影响。在这项研究中，我们推导出了组合多线圈数据的最佳方法，即使用信号与噪声平方的比值进行加权。我们表明，只要噪声是不相关的，这就是理论上的最佳组合。该方法在使用 7T 处的 32 通道接收线圈在四个不同脑区（左运动和右运动、枕叶皮层和内侧额皮层）中采集的体内质子 MRS 数据上进行了演示。
