论文部分内容阅读
医学成像过程复杂,且影响成像质量的因素非常多,导致医学影像质量的量化一直非常具有挑战性。信噪比是一种被广泛采用的医学影像质量量化方法,被定义为感兴趣区域的信号均值与背景噪声的标准差的商。它的计算和执行勾画的操作者及勾画的区域密切相关,因此有必要对其一致性进行系统的研究。文章共计使用了324例仿真的磁共振脑影像,分为正常脑和多发性硬化病变两类影像;并将脑脊液区域和白质区域作为感兴趣区域;最后采用Wilcoxon秩和检验来检验操作者间的差异性,并用Pearson相关系数和Spearman秩相关系数来验证感兴趣区域间的一致性。实验结果发现,对相同模态的相同组织区域,不同观察者间的信噪比值没有显著性差异(Wilcoxon秩和检验,P>0.70);对相同模态的相同观察者,Pearson相关系数rp>0.71(P<10~(-5)),而Spearman秩相关系数rs>0.97(P<10-3),即信噪比对磁共振影像质量的评价不受操作者以及感兴趣组织的选择而出现差异。进一步的临床数据验证有利于挖掘信噪比在临床应用中的使用条件和潜在局限。
The medical imaging process is complicated, and there are many factors that affect the imaging quality. As a result, the quantification of medical imaging quality has been very challenging. Signal-to-noise ratio is a widely used method of medical image quality quantification defined as the quotient of the signal mean of the region of interest and the background noise. Its calculation is closely related to the outlined operators and the outlined regions, so it is necessary to conduct a systematic study on its consistency. A total of 324 simulated magnetic resonance brain images were used and divided into two types of images: normal brain and multiple sclerosis. Cerebrospinal fluid and white matter regions were used as the region of interest. Wilcoxon rank-sum test Differences, Pearson correlation coefficient and Spearman rank correlation coefficient to verify the consistency between regions of interest. The experimental results showed that there was no significant difference in signal-to-noise ratio (Wilcoxon rank-sum test, P> 0.70) among different observers in the same region of the same modality. For the same observer with the same modalities, the Pearson correlation coefficient rp > 0.71 (P <10 -5), while Spearman rank correlation coefficient rs> 0.97 (P <10-3), that is, the evaluation of MR imaging quality was not affected by the operator and the tissue of interest The difference. Further validation of clinical data will help to explore the conditions of use and potential limitations of signal-to-noise ratio in clinical applications.