肾脏缺氧分数评价慢性肾病的初步研究

发布时间：2018-03-30 09:04

本文选题：肾脏缺氧分数　切入点：血氧水平依赖磁共振成像　出处：《苏州大学》2014年硕士论文

【摘要】：目的：探讨利用血氧水平依赖磁共振成像技术测算肾脏缺氧分数评价慢性肾病的价值。方法：本研究选取于2013年3月～2014年2月住院治疗的慢性肾病患者32例，分为两组（CKD1、2期为轻度组共16例，CKD3-5期为中重度组共16例）。并招募健康志愿者13例作为对照组。采用西门子3.0T磁共振进行常规T1WI、T2WI及mGRE序列BOLD-MRI扫描。由2名有经验的放射科医师各自进行2种不同的方法进行图像及数据计算、分析。方法一为传统方法，在Siemens Syngo图像分析软件上手工选取感兴趣区（ROI），完成数据测量并计算得到双肾皮质、髓质的R2*值。方法二为测算肾脏分数方法，选取整个肾脏为感兴趣区，利用spin软件，分析每个体素所对应的R2*值，计算R2*值35sec-1的体素所占全肾体素的百分比（即肾脏缺氧分数）。采用统计学计算分析(t检验、单因素方差分析、组内相关性分析、ROC曲线分析)，比较两种方法的诊断效能及可重复性。结果： 1.传统方法：正常对照组及慢性肾病组髓质R2*值均高于皮质。正常对照组与CKD轻、中重度组髓质R2*值有显著性差异（髓质F=13.634，P0.05），且R2*值随CKD严重程度加重逐渐增高。正常对照组与CKD轻度组间皮质R2*值无统计学差异（P0.05），正常对照组与CKD中重度组之间及CKD两组之间皮质R2*值有显著性差异，，中重度组明显大于轻度组和正常组。 2.肾脏缺氧分数方法：正常对照组与CKD轻度、中重度组的肾脏缺氧分数存在显著差异（F=60.144,P0.05），中重度组轻度组正常组。 3.利用ROC曲线，对比传统方法髓质R2*值与肾脏缺氧分数，肾脏缺氧分数的敏感性及特异性更高。 4.两种方法进行重复性分析，肾脏缺氧分数（重复系数=0.918）明显大于髓质R2*值（0.549），肾脏缺氧分数方法可重复性更好。结论：肾脏缺氧分数较传统手工选择ROI测量皮、髓质R2*值诊断敏感性及特异性更高，可重复性更高，更适合应用在CKD的BOLD-MRI图像分析中，可以更好的评价肾脏的氧化状态。
[Abstract]:Objective:. To evaluate the value of oxygen dependent magnetic resonance imaging (MRI) in the evaluation of chronic nephropathy. Methods:. From March 2013 to February 2014, 32 patients with chronic nephropathy were enrolled in this study. The patients were divided into two groups: mild group (n = 16), moderate and severe group (n = 16), and healthy volunteers (n = 13) as control group. Routine T1WI T2WI and mGRE sequence BOLD-MRI scanning were performed with Siemens 3.0T magnetic resonance imaging. Projectists perform two different methods for image and data calculation. Methods: method one is traditional method, the area of interest is selected manually on Siemens Syngo image analysis software, the data is measured and the R2 * value of the renal cortex and medulla is calculated, the second method is to calculate the kidney fraction, and to calculate the R2 * value of the renal cortex and medulla, and to calculate the R2 * value of the renal cortex and medulla. The whole kidney was selected as the region of interest. The R2 * value of each voxel was analyzed by spin software, and the percentage of R2 * 35sec-1 voxel to the whole renin was calculated. Single factor ANOVA, intra-group correlation analysis and ROC curve analysis were used to compare the diagnostic efficacy and repeatability of the two methods. Results:. 1. Traditional method: the medullary R2 * value of normal control group and chronic nephropathy group was higher than that of cortex. There was significant difference in medullary R2 * value between moderate and severe groups (medullary FN 13.634, P0.05, and R2 * increased gradually with the severity of CKD. There was no significant difference in R2 * value between normal control group and mild CKD group (P 0.05), and between normal control group and moderate and severe CKD group). There was significant difference in R2 * value of cortex between the two groups of CKD. The moderate to severe group was significantly larger than that of mild group and normal group. 2.Renal anoxia score: there was significant difference in renal hypoxia score between normal control group and mild CKD, moderate and severe group (P 0.05), and moderate and severe mild group (P 0.05). 3. The sensitivity and specificity of renal hypoxia fraction were higher than that of the traditional method. 3. ROC curve was used to compare the R2 * value of medulla with the anoxic fraction of kidney. 4. The repeatability of the two methods showed that the renal hypoxia fraction (repetition coefficient was 0.918) was significantly higher than the medullary R2 * value (0.549), and the renal hypoxia fraction method was more reproducible than that of the medullary R2 * value. Conclusion:. Compared with the traditional manual selection of ROI, the renal hypoxia fraction has higher diagnostic sensitivity and specificity, higher reproducibility and better evaluation of renal oxidation status in CKD BOLD-MRI image analysis.
【学位授予单位】：苏州大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：R692

【参考文献】