Kreis Roland
Departments of Radiology and Clinical Research, University of Bern, Bern, Switzerland.
Magn Reson Med. 2016 Jan;75(1):15-8. doi: 10.1002/mrm.25568. Epub 2015 Mar 6.
Cramér Rao Lower Bounds (CRLB) have become the standard for expression of uncertainties in quantitative MR spectroscopy. If properly interpreted as a lower threshold of the error associated with model fitting, and if the limits of its estimation are respected, CRLB are certainly a very valuable tool to give an idea of minimal uncertainties in magnetic resonance spectroscopy (MRS), although other sources of error may be larger. Unfortunately, it has also become standard practice to use relative CRLB expressed as a percentage of the presently estimated area or concentration value as unsupervised exclusion criterion for bad quality spectra. It is shown that such quality filtering with widely used threshold levels of 20% to 50% CRLB readily causes bias in the estimated mean concentrations of cohort data, leading to wrong or missed statistical findings--and if applied rigorously--to the failure of using MRS as a clinical instrument to diagnose disease characterized by low levels of metabolites. Instead, absolute CRLB in comparison to those of the normal group or CRLB in relation to normal metabolite levels may be more useful as quality criteria.
克拉美-罗下界（CRLB）已成为定量磁共振波谱中不确定性表达的标准。如果将其正确解释为与模型拟合相关误差的下限阈值，并且如果尊重其估计的限度，那么CRLB无疑是一个非常有价值的工具，可让人了解磁共振波谱（MRS）中的最小不确定性，尽管其他误差来源可能更大。不幸的是，将相对CRLB表示为当前估计面积或浓度值的百分比作为质量不佳谱图的无监督排除标准也已成为标准做法。结果表明，使用20%至50% CRLB的广泛使用的阈值水平进行此类质量过滤很容易导致队列数据估计平均浓度出现偏差，从而导致错误或遗漏的统计结果——如果严格应用——会导致无法将MRS用作诊断以低水平代谢物为特征疾病的临床工具。相反，与正常组的绝对CRLB或相对于正常代谢物水平的CRLB作为质量标准可能更有用。
