World Wide Web(WWW) services have grown to levels where significant delays are expected to happen.Technology like prefetching are likely to help users to personalize their needs ,reducing their waiting times. This pap...World Wide Web(WWW) services have grown to levels where significant delays are expected to happen.Technology like prefetching are likely to help users to personalize their needs ,reducing their waiting times. This paperfirstly describes the architecture of prefetching,then classifies them into three types: based on branch model,based ontree model and others and presents profoundly the basic ideas of some existing prefetching algorithms. Next,severalmodels for controlling the prefetching are introduced. At last ,the trend and course concerning the prefetching algo-rithms are concluded.展开更多
文摘目的:应用三维斑点追踪技术(three dimensional speckle tracking imaging,3D-STI)获取冠心病患者行经皮冠状动脉介入治疗(percutaneous coronary intervention,PCI)前后左室整体面积应变(global area strain,GAS)参数,并探讨面积应变参数预测PCI术后患者心功能改善的临床应用价值。方法:选取拟行PCI术的42例冠心病患者,分别于术前及术后6个月行常规三维超声心动图及3D-STI检查,记录左室舒张末期容积(end-diastolic volume,EDV)、左室收缩末期容积(end-systolic volume,ESV)、左室射血分数(left ventricular eject fraction,LVEF)、左室GAS值及PCI前后左室射血分数改变量(ΔLVEF)。根据PCI术后LVEF提高率是否≥5%将患者分为心功能改善组与未改善组,比较两组间各参数的差异并探讨面积应变参数与心功能改善间的相关性。结果:PCI术前与术后患者EDV,LVEF,ESV,LVEF,肌钙蛋白含量及GAS差异均无统计学意义(P>0.05)。与未改善组比较,改善组LVEF[(54.9±4.9)% vs (44.3±5.7)%,P<0.001],GAS[(?31.9±5.8)%vs (?26.0±5.9)%,P=0.003]均明显增高,但肌钙蛋白含量[(1.9±2.0)μg/L vs (4.0±3.4)μg/L]及ESV[(50.6±6.7) mL vs (63.2±10.9) mL]降低。冠心病患者PCI术前GAS与ΔLVEF呈正相关(r=0.58,P<0.001)。以术前GAS预测冠心病患者PCI术后心功能改善(LVEF提高率≥5%)的ROC曲线下面积为0.93,以GAS=28.9%为最佳截断点时敏感度为86.7%,特异度为88.9%。多元线性回归曲线结果显示:GAS(β=?0.73,P<0.001)可作为心功能改善的预测因素。结论:基于3D-STI的面积应变参数与PCI术后患者心功能改变相关,可用于预测冠心病患者PCI术后疗效。
文摘World Wide Web(WWW) services have grown to levels where significant delays are expected to happen.Technology like prefetching are likely to help users to personalize their needs ,reducing their waiting times. This paperfirstly describes the architecture of prefetching,then classifies them into three types: based on branch model,based ontree model and others and presents profoundly the basic ideas of some existing prefetching algorithms. Next,severalmodels for controlling the prefetching are introduced. At last ,the trend and course concerning the prefetching algo-rithms are concluded.
