摘要
目的探讨基于多层螺旋CT血管造影(MDCTA)测量的颈动脉直径和脑血管症状之间的关系,寻找与症状相关的颈动脉狭窄的阈值。方法以接受MDCTA检查的230例可疑颈动脉狭窄患者为研究对象,记录患者有无大脑前循环障碍症状。所有患者均行MDCTA检查测量狭窄颈动脉直径,同时观测动脉粥样硬化斑块类型。利用ROC曲线对症状和狭窄颈动脉直径进行分析确定诊断阈值,利用Logistic多元回归探讨症状与狭窄颈动脉直径、粥样硬化斑块类型和其他变量之间的关系。结果症状组颈动脉直径明显较无症状组小,ROC曲线分析发现曲线下面积为0.720,出现脑缺血症状的最佳预测值为1.6mm,灵敏度为0.734,特异度为0.638。Logistic回归分析证实颈动脉狭窄和软斑块是脑血管症状的独立危险因素。结论利用MDCTA测量颈动脉直径能够较为可靠地预测脑缺血症状,脑缺血症状与颈动脉狭窄和软斑块密切相关,这对于MDCTA在脑卒中防治领域进一步应用具有重要意义。
Objective To investigate the association between cerebral ischemia symptoms and carotid stenosis measured by multi-detector-row CT angiography (MDCTA) and to identify a reliable carotid stenosis threshold that was associated with cerebral ischemia symptoms. Methods 230 patients were studied for suspected carotid artery stenosis by using MDCTA and cerebral ischemia symptoms were recorded. In each patient, carotid artery stenosis was quantified and categories of atherosclerotic plaque were recorded using MDCTA. A ROC curve was calculated to identify a specific carotid stenosis threshold for the prediction of symptoms. Multivariate logistic regression analysis was used to estimate the strength of association between cerebral ischemia symptoms and carotid stenosis, categories of atherosclerotic plaque and other factors. Results Diameter of carotid artery in patients with symptoms was smaller than that of patients without symptoms. The ROC curve indicated that a threshold of 1.6 mm stenosis was associated with cerebral ischemia symptoms, with a sensitivity of 0. 734, a specificity of 0. 638, and AUC of 0. 720. Multiple logistic regression analysis confirmed that symptoms were associated with increased luminal stenosis and with the presence of soft plaques. Conclusion Luminal stenosis identified by MDCTA is a reliable predictor for cerebral ischemia symptoms. The results suggest an association between luminal stenosis, soft plaques and the presence of cerebral ischemia symptoms. MDCTA is worthy to apply further for prevention and treatment of stroke.
出处
《医学影像学杂志》
2012年第5期699-702,共4页
Journal of Medical Imaging