【摘要】 目的应用非参数逐步判别分析法(NSDA)建立颅内压(ICP)半定量数学模型,探讨其临床应用价值。方法对21例急性颅高压(ICH)病人采用硬膜外置探头监测ICP,同时连续20min采集大脑中动脉平均血流速度(VMCA)、平均动脉压(MAP)、呼气末二氧化碳分压(PETCO2)和心率(HR)。应用SAS软件,用这4个自变量建立3种ICP(正常至轻度升高、中度升高、重度升高)的NSDA半定量预测模型。结果模型对ICP正常至轻度升高、中度升高、重度升高阶段的错判率分别为5.20%、6.98%和10.17%,总体错判率为6.40%。结论NSDA模型对ICP的判定准确率较高,具有潜在临床应用前景。更多还原

【Abstract】 Objective To establish a semi-quantitative mathematical model of intracranial pressure(ICP) using nonparametric stepwise discriminant analysis(NSDA) and explore its clinical value. Methods ICP was monitored with an epidural sensor in 21 patients with acute intracranial hypertension(ICH) ,and mean blood flow velocity in the middle cerebral artery(VMCA) ,mean arterial pressure(MAP) ,end-tidal carbon dioxide partial pressure(PETCO2) and heart rate were collected continuously for 20 minutes at the same time. Semi-quantitative NSDA prediction models at three different ICP levels(normal or mildly elevated,moderately elevated and highly elevated) were established with the four independent variables by SAS software. Results The error rate of NSDA model for judging from normal ICP to the 3 elevated levels was 5.20%,6.98% and 10.17% respectively,and the total error rate was 6.40%. Conclusion NSDA model has high accuracy rate for ICP judgment,thus possessing potential perspective of clinical application.更多还原