经颅多普勒搏动指数预测脑白质病变程度的初步研究

Predicting the severity of white matter lesions with transcranial Doppler pulsatility index:a preliminary study

目的研究腔隙性脑梗死患者经颅多普勒(TCD)搏动指数(PI)与脑白质病变(WMLs)严重程度的关系。方法共纳入93例非心源性、不伴颅内外大动脉狭窄的腔隙性脑梗死患者。采用Fazekas评分,根据腔隙性脑梗死患者的头部MRI评估WMLs的严重程度,将患者分为无-轻度WMLs组(0～3分)和中-重度WMLs组(4～6分),分别为57例和36例。对全部患者行TCD检查,测量大脑中动脉的搏动指数(PI-MCA)及基底动脉的搏动指数(PI-BA)。应用Logistic回归评估中-重度WMLs的危险因素,并研究PI与WMLs严重程度的关系。结果①中-重度WMLs组的PI-MCA(1.10±0.21)和PI-BA(1.06±0.25)均高于无-轻度WMLs组的PI-MCA(0.91±0.13)和PI-BA(0.88±0.10),差异有统计学意义,P<0.001。②Logistic多元回归显示,PI-MCA、PI-BA、清同型半胱氨酸水平及既往卒中史,均为腔隙性脑梗死患者中-重度WMLs的独立预测因素。③受试者工作特征(ROC)曲线分析显示,PI-MCA>0.98时,预测中-重度WMLs的曲线下面积为0.811,敏感性为83.3%,特异性为78.9%,阳性预测率为71.4%,阴性预测率为88.2%。PI-BA≥0.95时,预测中-重度WMLs的曲线下面积为0.702,敏感性为66.7%,特异性为73.7%,阳性预测率为61.5%,阴性预测率为77.8%。结论 PI是反映脑白质病变严重程度的生理学指标。利用TCD检测MCA的PI值有助于预测腔隙性脑梗死患者脑白质病变的严重程度。

Objective To investigate the relation between the pulsatility index （PI） measured with transcranial Doppler （TCD） and the severity of white matter lesions （WMLs） in patients with lacunar infarction. Methods A total of 93 patients with lacunar infarction of non-cardiac origin and without extracranial-intracranial large artery stenosis were recruited. The patients were divided into either a non-or- mild WML group （0 -3 points, n =57） and a moderate-to-severe WML group （4 -6 points, n =36）. All the patients underwent TCD examination. Their PI of middle cerebral artery （PI-MCA） and PI of basilar artery （PI-BA） were measured. The logistic regression analysis was used to study the risk factors for mod- erate-to-severe WMLs and the relationship between PI and the severity of WMLs. Results （1） The PI-MCA （ 1.10 ± 0. 21 ） and PI-BA （ 1.06 ± 0. 25 ） in the moderate-to-severe WML group were higher than 0. 91±0. 13 and 0. 88±0. 10 in the non or mild WML group. There were significant differences （P 〈 0. 001 ） ; （2）Multiple logistic regression analysis showed that PI-MCA, PI-BA, serum homocysteine concen- tration and the history of previous stroke were the independent predictors for moderate-to-severe WMLs in patients with lacunar infarction. （3）Receiver operator curve （ROC） analysis showed that when PI-MCA was ≥ 0.98, the prediction of the area under the curve of moderate-to-severe WMLs was 0. 811, the sensitivity was 83.3% , specificity was 78.9% , positive predictive value was 71.4% , and negative predictive value was 88.2%. When PI-BA was ≥ 0.95, the prediction of the area under the curve of moderate-to-severe WMLs was 0. 702, the sensitivity was 66.7%, specificity was 73.7% , positive predictive value was 61. 5%, and negative predictive value was 77.8%. Conclusion PI is a physiological index of reflecting the severity of WMLs. Using TCD to detect the PI value of MCA contributes to predict the severity of white matter lesions in patients with lacunar infarction.