痉挛性脑性瘫痪早产儿脑CT图像形态学危险因素的相关分析  

Correlation analysis of CT imaging morphological risk factors in prematures with spastic cerebral palsy

作　　者：李鹤虹[1] 柳国胜 罗良平 

机构地区：[1]广州市儿童医院CT室,广东省广州市510120 [2]暨南大学华侨医院儿科,广东省广州市510120 [3]暨南大学华侨医院放射科,广东省广州市510120

出　　处：《中国临床康复》2005年第27期166-169,共4页Chinese Journal of Clinical Rehabilitation

摘　　要：目的:对痉挛性脑性瘫痪早产儿的断层影像资料进行相关因素分析,以确定早产儿痉挛性脑性瘫痪的形态学危险因素的图像特征。方法:①选择1997-10/2004-06广州市儿童医院新生儿科住院治疗的新生儿87例,男58例,女29例;平均年龄1岁11个月,男孩年龄5个月～6岁,女孩年龄6个月～5岁1个月。孕龄均小于36周;均有完整CT图像资料;监护人知情同意。②分析87例脑瘫患儿的CT影像资料,按形态学相关因素分为5组:X1(脑萎缩);X2(脑室扩大);X3(灰质异常增多异位);X4(白质丢失);X5(脑室周围白质角度)。③采用SPSS11.0统计软件对早产儿痉挛性脑瘫的形态学相关因素进行非条件logistic单因素及多因素回归分析,第一步将X4(白质丢失)作自变量强迫引入,Y作应变量。第2步将X1,X2,X3,X5变量采用向后剔除似然法由计算机模型逐步筛选,对肢体瘫痪变量的危险因素暴露度,回归分析结果作假设检验确定回归方程的意义。结果:①Logistic单因素及多因素回归分析结果表明侧脑室容积大小对于早产儿痉挛性脑瘫危险因素不如灰质异常,其危险度(诊断意义)及特异度显然不能与灰质异常增多和异位相比(多因素分析:脑萎缩危险因素的OR<0.7,P=0.032,wald值为4.624;脑室扩大变量的OR=0.752,P=0.536,wald值为0.383;而灰质异常增多和异位的OR=8.874,P=0.013,偏回归系数为2.183,wald值为6.142;白质丢失OR=1.792,P=0.039,偏回归系数为-0.233,wald值为4.274,95%CI:0.525～2.494;室周白质角度OR=4.311,偏回归系数为1.461,wald值为0.398,95%CI:1.079～17.227)。灰质异常增多和异位、室周白质角度及白质丢失是肢体痉挛脑性瘫痪的危险因素,其中灰质异常增多和异位较白质丢失和室周白质角度更为重要。②脑萎缩和脑室扩大对判断肢体痉挛脑瘫无特异性,年龄因素亦被回归模型排除。结论:早产儿脑灰质异位、增多等异常是痉挛性脑性瘫痪的主要相关因素,此AIM： To analyze the CT imaging data of 87 prematures with spastic cerebral palsy, so as to identify the morphological risk factors and image characteristics in s prematures with spastic cerebral palsy. METHODS： ①Eighty-seven neonates at an average age of 23 months, including 58 boys （aged 5 months to 6 years） and 29 girls （aged 6 months to 5 years and one month）, who were Hospitalized in the Department of Neonatology, Guangzhou Children＇s Hospital between October 1997 and June 2004, were involved in the study with the permission of their guardians. Their gestational ages were all less than 36 weeks, and they all had complete CT imaging data. ②The CT imaging data of 87 children with spastic cerebral palsy were divided into 5 groups according to the morphological related factors： X1 （cortical atrophy）, X2 （dilation of cerebral ventricles）, X3 （abnormal gray matter, included abnormal increment and heterotopia）, X4 （white matter losing） and X5 （angle of periventrieular white matter）, ③The morphological related factors of prematures with spastic cerebral palsy were analyzed with non-conditional logistic univariate and multivariate regression analysis by using the SPSS 11.0 software： Firstly, X4 （white matter losing） was introduced compulsively as the independent variable, and Y as the dependent variable; Secondly, the variables of X1,X2, X3 and X5 were screened step by step with computer model by means of backward remove likelihood, the hypothesis testing was performed to the exposed degree o the risk factors of the variables of limb palsy and the results of the regression analysis so as to identify the significance of the regression equation. RESULTS： ① The results of Logistic univariate and multivariate regression analysis showed that the influence of the risk factor of cerebral ventrieular volume on prematures with spastic cerebral palsy was not as much as that of abnormal gray matter, its risk degree （diagnostic significance） and specificity were remarkably n

关 键 词：婴儿 早产 脑性瘫痪 回归分析 体层摄影术 X线计算机 

分 类 号：R742.3[医药卫生—神经病学与精神病学]