摘要
目的通过对彩色多普勒血流显像(CDFI)检查颈内动脉(ICA)颅外段狭窄相关的血流动力学测量参数进行判别和主成分分析,完善判别方法。方法以华扬等提出的测量参数为标准,对86例ICA轻度狭窄和93例ICA中、重度狭窄患者进行CDFI常规检测,ICA中、重度狭窄患者均经全脑数字减影血管造影(DSA)诊断证实。应用判别分析(Fisher差别法)和主成分分析方法对CDFI测量的颈内动脉狭窄段收缩期峰值流速(PSVICA)、颈内动脉狭窄段舒张期峰值流速(EDVICA)和PSVICA/颈内动脉狭窄远段收缩期峰值流速(PSVDIS)数值进行统计学分析。并推导函数式。结果①轻、中、重度狭窄的Fisher线性判别函数式分别为:Y1=0.105×PSVICA+0.142×EDVICA+1.247×PSVICA/PSVDIS-10.769;Y2=0.022×PSVICA+0.411×EDVICA+12.552×PSVICA/PSVDIS-36.773;Y3=0.145×PSVICA+0.560×EDVICA+14.018×PSVICA/PSVDIS-88.392。②由主成分分析得到的轻、中、重度狭窄得分综合评价函数式分别为:f1=0.017291×PSVICA+0.016535×EDVICA+3.626682×PSVICA/PSVDIS-7.225329;f2=0.005747×PSVICA+0.040419×EDVICA+0.506257×PSVICA/PSVDIS-5.675821;f3=0.006775×PSVICA+0.030777×EDVICA+0.446399×PSVICA/PSVDIS-7.842303。③主成分分析判别狭窄程度临界值,轻度以下为f1<1.89;中度为f1>1.89,同时f3<-1.7;重度为f3>-1.71。临界值判断与DSA的符合率达98.9%。结论不断完善的CDFI检测方法是检查ICA狭窄重要、快捷的手段。其中采用主成分分析,利用临界值判断,与DSA的符合率很高。
Objective To perform discriminant analysis and principal component analysis for extracranial internal carotid artery （ICA） stenosis-associated hemodynamic parameters through color Doppler flow image （CDFI） and to perfect the discrimination method. Methods Eighty-six patients with mild ICA and 93 patients with moderate to severe ICA stenosis confirmed by global digital subtraction arteriography （DSA） were examined by conventional CDFI according to the measuring parameters presented by HUA Yang et al as criteria. The statistical analysis were performed according to the peak systolic velocity in ICA （PSVICA）, end diastolic velocity in ICA （EDVICA ） and PSVICA/Peak systolic velocity in distal internal carotid stenosis（PSVDIS）. Results The mild, moderate and severe Fisher＇s linear discriminant function were as follows： Y1 =0. 105×PSVICA ＋0. 142×EDVICA ＋ 1. 247×PSVICA/PSVDIS - 10. 769; Y2 =0. 022×PSVICA ＋ 0. 411×EDVICA ＋ 12. 552×PSVICA/PSVDIS - 36. 773 ; and Y3 = 0. 145×PSVICA ＋ 0. 560×EDVICA ＋ 14. 018×PSVICA/ PSVDIS -88. 392. The mild, moderate and severe comprehensive evaluation function of principle components scores obtained from the principal component analysis were as follows ： f1 = 0. 017291×PSVICA ＋ 0. 016535×EDVICA ＋ 3. 626682×PSVICA/VSVDIS - 7. 225329; f2 = 0. 005747×PSVICA ＋0. 040419×EDVICA ＋ 0. 506257×PSVICA/PSVDIS - 5. 675821 ; and f3 = 0. 006775×PSVICA ＋ 0. 030777×EDVICA ＋ 0. 446399×PSVICA/PSVDIS - 7. 842303. The estimated stenotie degree of critical values were mild f1 〈 1.89, moderate f1 〉 1.89, f3 〈 - 1.71 ,and severe f3 〉 - 1.71. The coincidence rate of critical value estimation and DSA reached 98.9%. Conclusion The constantly perfected CDFI method is an important and shortcut approach in detecting ICA stenosis. Among them, using the principle component analysis, and critical value estimation have a high coincidence rate with DSA.
出处
《中国脑血管病杂志》
CAS
2008年第6期246-250,共5页
Chinese Journal of Cerebrovascular Diseases
关键词
超声检查
多普勒
彩色
颈动脉狭窄
判别分析
主成分分析
Ultrasonography, Doppler, Color
Carotid stenosis
Discriminant analysis
Principle component analysis
