经颅多普勒超声诊断大脑中动脉狭窄准确性的研究

目的评估经颅多普勒超声（TCD）的血流动力学不同参数对大脑中动脉（MCA）轻（〈50％）、中（50％～69％）及重度（70％～99％）狭窄诊断的准确性。方法连续纳入临床疑诊为MCA狭窄患者298例（523支MCA，剔除73支闭塞的血管）。行TCD检查，并记录MCA狭窄段，狭窄近、远段的收缩期峰值流速（PSVst、PSVpro、PSVpro）及狭窄段平均流速（MFVst）；计算PSVst／PSVpro、PSVst／PSVpro。以DSA结果为标准，通过受试者工作特征曲线（ROC曲线）下面积，确定诊断不同程度MCA狭窄的最佳参数界值，并分析不同参数诊断MCA狭窄的敏感性、特异性、准确性、阳性预测值及阴性预测值。结果①DSA显示。523支MCA中，正常血管及轻、中、重度狭窄的血管分别为168、61、71及223支。②以140cm／s≤PSVst〈180cm／s和90cm／s≤MFVst〈120cm／s作为诊断MCA轻度狭窄的界值时，准确性最高，分别为87．3％（200／229支）和84．7％（194／229支）。③以180cm／s≤PSVst〈220cm／s、120cm／s≤MFVst〈150cm／s、2．0≤PSVst／PSVpro〈3．0、1．8≤PSVs/PSVpro〈2．0作为诊断MCA中度狭窄界值时，准确性最高，依次为89．0％（267／300支）、87．7％（263／300支）、87．7％（263／300支）、85．7％（253／300支）。④以PSVst≥220cm／s、MFVst≥150cm／s、PSVst／PSVpro≥3．0及PSVst／PSVpro≥2．0作为诊断MCA重度狭窄界值时，诊断准确性最高，依次为88．0％（460／523支）、87．4％（457／523支）、88．9％（465／523支）及83．9％（439／523支）。与上述参数比较，联合PSVst≥220cm／s和PSVst／PSVpro≥3．0诊断MCA重度狭窄，可提高诊断的特异性（92．3％）。结论以DSA为参照标准，TCD诊断不同程度MCA狭窄的准确性较高。联合PSYst和PSV／PSYpro是鉴别MCA重度狭窄的重要指标。

The diagnostic accuracy of transcranial Doppler ultrasonography for middle cerebral artery stenosis

Objective To assess the diagnostic accuracies of the different hemodynamic parameters of transcranial Doppler uhrasonography （TCD） for mild （ 〈 50% ）, moderate （50-69%）, and severe （70-99%） stenosis of middle cerebral artery （MCA）. Methods 523 MCA in 298 consecutive patients （73 occluded MCA were excluded） were clinically suspected of having MCA stenosis. TCD examination was performed, the systolic peak velocity （PSVst, PSVpro, PSVpro） of the stenotic segments of MCA and the proximal and distal segments of the stenosis, as well as the mean flow velocity （MFVst） of the stenotic segments were recorded. The PSVst/PSVpro and PSVst/PSVpro were calculated. Taking the DSA results as the standard, the optimal parameter for diagnosing different degrees of MCA stenosis were determine by the receiver operating characteristic curve （ ROC curve）, and the sensitivity, specificity, accuracy, positive, and negative predictive value for diagnosing MCA stenosis with different parameters were analyzed. Results ① Of the 523 MCA showed by DSA, the normal, mild, moderate, and severe stenosis were 168, 61, 71, and 223, respectively.② When taking 140 cm/s ≤PSVst 〈 180 cm/s and 90 cm/s ≤MFVst 〈 120 cm/s as the boundary values of diagnosing mild MCA stenosis, the accuracies were the highest, and they were 87.3% （200/229 branches） and 84. 7% （194/229 branches） respectively. ③ When taking 180 cm/s ≤ PSVst 〈220 cm/s, 120 cm/s ≤MFVst 〈 150 cm/s, 2.0 ≤PSVst/PSVpro 〈3.0, and 1.8 ≤PSVst/ PSVpro 〈 2. 0 as the diagnostic boundary values of diagnosing moderate MCA stenosis, the accuracies were the highest, and they were 89.0% （267/300 branches）, 87.7% （263/300 branches）, 87.7% （263/ 300 branches）, and 85.7% （253/300 branches）, respectively. ④ When taking PSVst ≥ 220 cn/s, MFVst ≥ 150 cm/s, PSVst/PSVpro ≥3.0, and PSVst/PSVpro ≥2. 0 as the diagnostic boundary values of diagnosing severe MCA stenosis, the accuracies were highest, and they were 88.0% （460/523 branches） , 87. 4% （457/523 branches）, 88.9% （465/523 branches）, and 83. 9% （439/523 branches）, respectively. As compared to the above parameters, the diagnosis of severe MCA stenosis in combination of PSVst ≥220 cm/s and PSVst/PSVpro ≥ 3.0 could increase the specificity of diagnosis （92.3%）. Conclusions Take the DSA findings as the standard, the accuracies of diagnosing different degrees of MCA stenosis with TCD is higher. Tile combination of PSVst and PSV/PSVpro is the important criterion for identifying severe MCA stenosis.