常规MRI用于总成熟度评分量化早产儿脑成熟度及其预后的价值

MRI quantitative assessment of brain maturation and prognosis in premature infants using total maturation score

目的探讨常规MRI总成熟度评分（TMS）在量化早产儿脑成熟度及其预后的价值。方法搜集复查头部MRI有后遗脑白质损伤（WMD）的早产儿19例（后遗WMD组）和无脑损伤早产儿21名（对照组）。由2名医师对2组早产儿围产期的MRI进行TMS，包括髓鞘化、皮层折叠、胚胎生发基质和神经胶质细胞移行带评分。使用组内相关系数（ICC）评价2名医师诊断的信度、一致性。应用简单线性回归分析早产儿TMS评分与其修正月龄（PMA）的相关性，获得直线回归方程，并对2组评分进行独立样本t检验。结果后遗WMD组19例患儿围产期有16例MRI示半卵圆中心、侧脑室旁点簇状或线状的短T1、短T2信号。DWI除上述病灶呈高信号外，3例出现原病灶范围增大，4例在胼胝体膝部、压部出现新的片状高信号灶。对照组21名健康早产儿围产期MRI及DWI均未见异常信号影。2名医师TMS评分分别为（7．13±2．27）、（7．13±2．21）分，观察者间信度较高（ICC=0．990，P〈0．05）。后遗WMD组与对照组的TMS与围产期MR检查时的PMA均呈正相关（r2值分别为0．6401、0．5156，P值均〈0．05），回归方程为Y：0．733X-19．421，Y=0．846X-21．835（Y：TMS，X：PMA）。后遗WMD组与对照组TMS评分分别为（6．32±1．73）、（7．86±2．48）分，其中2组的髓鞘化评分分别为（1．95±0．71）、（2．52±0．75）分，后遗WMD组较对照组髓鞘化显著延迟；皮层折叠评分分别为（2．90±1．09）、（2．26±0．87）分，差异均有统计学意义（t值分别为2．26、2．50、2．04，P值均〈0．05）。胚胎生发基质评分分别为（1．00±0．01）、（1．14±0．36）分，神经胶质细胞移行带评分分别为（1．11±0．46）、（1．29±0．78）分，差异均无统计学意义（t值分别为1．83、0．88，P值均〉0．05）。结论常规MRI借助TMS评分能半定量评价早产儿脑成熟度，预测早产儿的WMD。

Objective To quantitatively assess brain maturation and prognosis in premature infants on conventional MRI using total maturation score （TMS）. Methods Nineteen cases of sequelae of white matter damage （ WMD group ） and 21 cases of matched controls （ control group ） in premature infants confirmed by MRI examinations were included in the study. All cases underwent conventional MR imaging approximately during the perinatal period after birth. Brain development was quantitatively assessed using Childs AM＇s validated scoring system of TMS by two sophisticated radiology physicians. Interobserver agreement and reliability was＇ evaluated by using intraclass correlation （ICC）. Linear regression analysis between TMS and postmenstrual age （PMA） was made（ Y： TMS, X： PMA）. Independent-sample t test of the two groups＇ TMS was made. Results Sixteen of 19 cases revealed MR[ abnormalities. Lesions showing Tl and T2 shortening tended to occur in clusters or a linear pattern in the deep white matter of the centrum semiovale, periventricular white matter. Diffusion-weighted MR image （DWI） showed 3 cases with greater lesions and 4 cases with new lesions in corpus callosum. There was no abnormality in control group on MRI and DWI. The average numbers of TMS between the two observers were 7.13 ＋ 2. 27, 7. 13 ＋ 2. 21. Interobserver agreement was found to be high （ICC =0. 990, P 〈0. 05 ） . Positive correlation was demonstrated between TMS and PMA in MRI（r2 = 0. 6401,0. 5156 respectively, P 〈 0.01 ） of the WMD group and the control group. Linear regression equation showing positive correlation between TMS and PMA were Y = 0. 733X - 19. 421, Y = 0. 846X - 21. 835 （ P 〈 0. 05 ）. TMS of the two groups （ average numbers are 6. 32 ＋ 1.73, 7.86 ＋2.48 respectively） were statistically significantly different （t =2. 26, P 〈0. 05）. Myelination （average numbers are 1.95 ＋ 0. 71,2. 52 ＋ 0. 75 ） appeared to be significantly different（ t = 2.50, P 〈 0.05 ）, which delayed in babies with SWMD group. Cortical infolding （average numbers are 2. 26 ＋ 0. 87, 2. 90 ＋ 1.09 ） appeared to be different （t = 2.04, P 〈 0. 05 ） without clinical significance. Germinal matrix distribution （average numbers are 1.00 ＋0. 01, 1.14＋ 0. 36 respectively） and bands of migrating glial cells （ 1.11 ＋0. 46, 1.29 ＋0. 78 respectively） showed no statistical difference between WMD group and control group （ t = 1.83, t = 0. 88, P 〉 0. 05 ）. Conclusion Conventional MRI is able to quantify the brain maturation and prognosis of premature infants using TMS.