论文部分内容阅读
目的比较不同CT灌注后处理技术所预测的缺血性病灶,并将CT灌注确定的病灶与卒中病人最终病灶的大小相比较。方法 50例病人进行CT、CT灌注和CT血管成像。采用最小均方去卷积(LMSD)、最大斜率(MS)和传统的奇值分解去卷积算法(SVDD)计算出各量化值与伪彩图。比较这些算法之间各定量值、核心/半暗带大小和Alberta卒中项目早期CT(ASPECTS)评分;病灶大小和ASPECTS评分与随访的MRI+MRA或者CT+CTA中的最终病灶为参考标准进行比较,评估血管再通的状态。结果各量化值与病灶大小之间存在统计学差异,但是在所有病例中根据ASPECTS评分和核心/半暗带比例所做的治疗计划都是一致的。CT灌注成像中的病灶大小能有效预测最终的梗死范围:对于血管再通组,采用LMSD、MS和SVDD算法时CT灌注与随访病灶大小的一致性系数分别为0.87、0.82和0.61(P<0.001);而对于非血管再通组,三者则分别为0.88、0.87和0.76(P<0.001)。结论 CT灌注像中的病灶能够有效预测最终的梗死灶大小。不同的CT灌注后处理算法常采取相同的临床治疗决定,但是对于病灶大小的评价,LMSD、MS算法要优于SVDD算法。原文载于EurRadiol,2012,22(12):2559-2567.
Objective To compare ischemic lesions predicted by different CT perfusion postprocessing techniques and to compare the final lesion sizes determined by CT perfusion with those of stroke patients. Methods Fifty patients underwent CT, CT perfusion and CT angiography. The quantized values and the pseudo-color map were calculated by least mean square deconvolution (LMSD), maximum slope (MS) and traditional singular value decomposition deconvolution algorithm (SVDD). Comparisons of the quantitative, core / penumbra size, and early-stage CT (ASPECTS) scores of these algorithms between the algorithms were performed; the lesion size and ASPECTS scores were compared with the final lesion in follow-up MRI + MRA or CT + CTA as a reference standard , Assess the status of revascularization. Results There was a statistically significant difference between the quantified values and the size of the lesions, but the treatment plan was based on the ASPECTS score and core / penumbra ratio in all cases. The size of the lesion in CT perfusion imaging effectively predicted the final infarct size. For the recanalization group, the coefficients of agreement between CT perfusion and follow-up lesion size were 0.87, 0.82, and 0.61 (P <0.001, respectively) using the LMSD, MS, and SVDD algorithms ), Whereas for the non-revascularization group, the three were 0.88, 0.87 and 0.76, respectively (P <0.001). Conclusion CT perfusion in the lesion can effectively predict the final size of the infarct. Different CT perfusion postprocessing algorithms often take the same clinical treatment decisions, but for the size of the lesion, LMSD, MS algorithm is better than the SVDD algorithm. The original is contained in EurRadiol, 2012, 22 (12): 2559-2567.