摘要
目的探讨CT能谱成像技术预测肝硬化门静脉高压患者并发食管静脉曲张破裂出血(EVB)风险的价值。方法回顾性分析肝硬化门脉高压患者60例(其中30例并发EVB,30例无EVB)及肝脏正常的30例患者资料。绘制ROC曲线,分别评价根据门静脉期肝左、右及尾状叶的碘浓度和胃冠状静脉(GCV)直径预测EVB的诊断效能。选取敏感度最高的肝叶碘浓度与GCV直径联合行判别分析,评价其联合诊断价值。结果当肝左、右及尾状叶的碘浓度及GCV直径预测EVB临界值分别为17.56mg/ml、18.29mg/ml、23.01mg/ml、5.8mm时,曲线下面积分别为0.833、0.874、0.701,0.726,敏感度及特异度分别为80.0%、86.7%、70.0%、73.3%和66.7%、76.7%、60.0%、63.3%,阳性预测值及阴性预测值分别为76.9%、85.2%、62.1%、66.7%和70.6%、78.8%、63.6%、69.2%。联合肝右叶碘浓度与GCV直径判别分析提示样本回代符合率为88.3%。结论 CT能谱成像技术可用于预测肝硬化门静脉高压患者并发EVB的风险。
Objective To observe the value of spectral CT imaging on predicting the risk of esophageal variceal bleeding （EVB） in portal hypertension patients caused by liver cirrhosis. Methods Spectral CT imaging data of 60 patients with liv- er cirrhosis and portal hypertension （30 patients complicated with EVB, 30 patients without EVB） and 30 patients with normal liver were analyzed retrospectively. ROC curve was drawn to evaluate the diagnosis performance that use iodine con- centrations of left, right and caudate lobe of liver in portal phase and diameter of gastric coronary vein （GCV） to predict EVB. The highest sensitivity and the diameter of GCV were selected to evaluate the joint diagnostic value that the iodine concentration of liver lobe. Results When cut-off value of iodine concentrations in left, right and caudate lobe of liver and the diameter of GCV was 17. 56 mg/ml, 18. 29 mg/ml, 23.01 mg/ml, 5.8 ram, respectively. The area under ROC curve was 0. 833, 0. 874, 0. 701, 0. 726, respectively. The sensitivity and specificity was 80. 0%, 86.7%, 70. 0%, 73.3% and 66.7%, 76.7%, 60. 0%, 63.3%, respectively, the positive predictive value and negative predictive value was 76.9%, 85.2%, 62.1%, 66.7% and 70.6%, 78.8%, 63.6%, 69.2%, respectively. The coincidence rate of the sample back was 88.3% that joint the concentration of liver right lobe and GCV diameter. Conclusion Spectral CT imaging can be used to predict the risk of EVB for patients with liver cirrhosis and portal hypertension.
出处
《中国医学影像技术》
CSCD
北大核心
2012年第12期2201-2205,共5页
Chinese Journal of Medical Imaging Technology
