CTA源图像上CT值变化率预测急性脑梗死后脑实质出血:基于ASPECTS软件脑梗死灶评估

CT value change rate of cerebral infarction on CT angiographic source images can predict cerebral parenchymal hemorrhage after acute ischemic stroke:a study based on ASPECTS software for evaluation the cerebral infarction

目的:探讨CTA源图像(CTASI)上缺血脑组织CT值变化率结合多种临床和影像学指标预测急性缺血性卒中患者发生脑实质出血(PH)的价值。方法:回顾性搜集本院2016-2020年因前循环大血管闭塞导致急性缺血性卒中的172例患者的临床和影像资料。所有患者行CT平扫和CTA检查。利用自动化Alberta卒中项目早期CT评分(ASPECTS)软件,分别在平扫CT和CTASI图像上确定梗死灶的范围,测量并计算所有梗死区(AvCT_(梗死))和对侧非梗死区的平均CT值(AvCT_(非梗死)),基于两者的比值计算梗死灶的CT值变化率。将美国国立卫生研究院卒中量表(NIHSS)评分、卒中发作距CT检查的间隔时间、ASPECTS、血管闭塞位置、梗死部位和侧枝代偿情况等变量纳入logistic回归分析,筛选PH的预测因素,并采用ROC曲线分析其预测效能。当血肿面积>梗死面积的30%时定义为PH2。结果:CTASI上梗死灶的CT值变化率(OR=6.505、5.944;P均<0.05)及患者的NIHSS评分(OR=1.067、1.104;P均<0.05)为PH和PH2的独立预测因素。预测PH时,CTASI上梗死灶CT值变化率的AUC为0.79;CTASI上测量的CT值变化率联合NIHSS评分预测PH的AUC为0.81,相应敏感度为85.71%。预测PH2时,CTASI上CT值变化率的AUC为0.78,联合NIHSS评分后AUC提高至0.84,相应敏感度为92.31%。结论:CTASI上梗死灶的CT值变化率和NIHSS评分是急性缺血性卒中患者发生PH的独立预测因素,两者结合能够提高预测效能,尤其是预测的敏感度。

Objective:The purpose of this study was to evaluate the CT value change rate of infarcted brain tissue on CTA source images(CTASI)combined with multiple clinical and imaging features to predict the occurrence of cerebral parenchymal hemorrhage(PH)in patients with acute ischemic stroke.Methods:A retrospective collection of the clinical and imaging data of 172 patients with acute ischemic stroke due to large vessel occlusion in the anterior circulation in our institution from 2016 to 2020 were performed.All subject underwent non-enhanced CT and CTA examination.ASPECTS areas with brain infarction(BI)were identified on non-enhanced CT and CTASI images using the Alberta Stroke Program Early CT Score(ASPECTS)software,and the average CT values of all the ASPECTS areas with BI and all the contralateral ASPECTS areas without BI were measured and calculated and named as AvCTIBT and AvCTnon-BI respectively,and then the CT value change rate of BI areas were calculated according to the ratio of AvCT_(BI)/AvCT_(non-BI).Variables such as National Institutes of Health Stroke Scale score(NIHSS),time from stroke onset to CT scan,ASPECTS,location of vascular occlusion,site of infarction,and collateral compensation were analyzed using logistic regression analysis and ROC analysis for selecting the predictors of PH and evaluating their predictive efficacy.PH2 was defined when the haematoma area was greater than 30%of the infarct area.Results:The CT value change rate on CTASI(OR=6.505 and 5.944,respectively;both P<0.05)and NIHSS score(OR=1.067 and 1.104,respectively;both P<0.05)were independent predictors of PH and PH2.For predicting PH,the AUC of the CT value change rate on CTASI were 0.79;and the AUC of the CT value change rate on CTASI combined with the NIHSS score was 0.81,with sensitivity of 85.71%.For predicting PH2,The AUC of the CT value change rate on CTASI was 0.78,and when combined with NIHSS score,the AUC was improved to 0.84 with sensitivity of 92.31%.Conclusion:The CT value change rate on CTASI and NIHSS are independent predictors for identifying patients with acute anterior circulation large vessel occlusive ischemic stroke who develop PH,and the combination of both predictors can improve predictive efficacy and sensitivity.