MRI灰度直方图分析在髓母细胞瘤复发风险评估中的应用

发布时间：2018-04-24 16:23

本文选题：髓母细胞瘤 + 复发　；参考：《中国介入影像与治疗学》2017年08期

【摘要】：目的探讨MRI灰度直方图分析在髓母细胞瘤复发风险评估中的应用价值。方法回顾性分析我院经手术病理证实的28例后颅窝髓母细胞瘤患者资料,根据术后随访3年内肿瘤是否复发将其分为复发组及未复发组,各14例。选取肿瘤MR增强矢状位图像最大层面,采用Mazda软件勾画ROI并进行灰度直方图分析,对2组肿瘤直方图参数特征进行统计学分析,获得2组肿瘤间差异有统计学意义的参数特征,并绘制ROC曲线,评价其对髓母细胞瘤复发的诊断效能。结果利用直方图提取出的9个参数特征中,2组间峰度的差异有统计学意义(P=0.018),其ROC曲线下面积为0.776(P=0.018),最佳临界值为1.268,峰度鉴别肿瘤复发风险的敏感度、特异度分别为64.3%、71.4%。结论 MRI灰度直方图分析可作为评估髓母细胞瘤复发风险的重要辅助手段。
[Abstract]:Objective to evaluate the value of MRI gray histogram analysis in assessing the recurrence risk of medulloblastoma. Methods the data of 28 cases of posterior fossa medulloblastoma confirmed by operation and pathology in our hospital were retrospectively analyzed. According to whether the tumor recurred within 3 years after operation, they were divided into recurrent group and non-recurrence group, 14 cases each. Selecting the largest layer of Mr enhanced sagittal image of tumor, using Mazda software to sketch ROI and carry out gray histogram analysis, the parameter characteristics of tumor histogram in two groups were statistically analyzed, and the difference between two groups of tumor parameters was obtained. ROC curves were plotted to evaluate the diagnostic efficacy of ROC in the diagnosis of medulloblastoma recurrence. Results there was significant difference in kurtosis between the two groups among the nine parameters extracted by histogram. The area under the ROC curve was 0.776 ~ (-1) ~ (-1) ~ (-1), and the best critical value was 1.268. The sensitivity of kurtosis in identifying the risk of tumor recurrence was 64.3% and the specificity was 71.4% respectively. Conclusion MRI grayscale histogram analysis can be used to evaluate the recurrence risk of medulloblastoma.
【作者单位】：郑州大学第一附属医院磁共振科;
【分类号】：R445.2;R739.41

【参考文献】