摘要
目的:探讨肿瘤周围区域的磁共振波谱在星形细胞肿瘤恶性程度分级中的作用。材料和方法:51例病理证实的星形细胞肿瘤患者分为低度恶性(WHOⅠ~Ⅱ级)和高度恶性(WHOⅢ~Ⅳ级),所有病例术前均行二维多体素磁共振波谱检查,测量肿瘤周围区域代谢物的相对定量,进行统计学比较,同时获得ROC曲线。结果:高度和低度恶性星形细胞肿瘤的瘤周相同部位的比较显示在瘤周1、2区域Cho/Cr、Cho/NAA、NAA/Cr之间有统计学差异,其中CNI(Cho/NAA index)在两个体素区域均有显著性差异(P<0.01);CNI在低度恶性星形细胞肿瘤的瘤周体素1、2依次为0.897±0.063、0.726±0.055,在高度恶性星形细胞肿瘤的瘤周体素1、2依次为1.566±0.113、1.131±0.079;利用ROC曲线显示瘤周体素1区域CNI 1.102时,AUC为0.86,敏感性和特异性分别为81.3%和81.6%;在瘤周体素2区域CNI0.925时,AUC为0.815,敏感性和特异性分别为70.4%和87.5%。结论:瘤周区域的代谢变化反映了星形细胞肿瘤对于肿瘤周围的浸润程度,从而对肿瘤的恶性度分级有帮助。
Purpose: Objective to evaluate the value of MR spectrum of the peritumoral region in the grading of intracranial astrocytic tumors. Materials and Methods: Fifty - one cases of intracranial astrocytic tumors were underwent the MR spectroscopy examination including 24 cases of low - grade astrocytic tumors and 27 cases of high - grade astrocytic tumors. The metabolic changes of the peritumoral region were measured by the relative quantification and comparison was made between the low and high grade astrocytic tumors. ROC(receive operating characteristic curve) was obtained at the same time. Results: Statistical significance of Cho/Cr,Cho/NAA、NAA/Cr can be seen in the peritumoral region between the low and high - grade astrocytic tumors. Remarkable significance(P 〈 0.01) of CNI (Cho/ NAA index)can be seen in the adjacent voxel 1 and the voxel 2 away from the tumor in the peritumoral region, it was 0.897 ± 0. 063、0. 726 ± 0.055 in the low - grade astrocytic tumors and 1. 566 ± 0. 113、 1. 131 ± 0. 079 in the high - grade astrocytic tumors. According to the ROC, sensitivity and specificity is 81.3 and 81.6% as CNI is 1. 102 in the voxel 1 and is 70.4% and 87.5% as CNI is 0.925 in the voxel 2. While AUC(area under curve) is 0.86 and 0. 815 respectively. Conclusion: The metabolic changes in the peritumoral region reflect the infiltration of the astrocytic tumors and can be helpful in the malignancy classification.
出处
《中国医学计算机成像杂志》
CSCD
2008年第4期281-286,共6页
Chinese Computed Medical Imaging
关键词
磁共振成像
波谱成像
星形细胞肿瘤
恶性
分级
Magnetic resonance imaging
MR spectroscopy
Astrocytic tumor
Malignant
Classification