Pap A, Boda K
First Department of Medicine, University Medical School, Szeged, Hungary.
Int J Pancreatol. 1986 Oct;1(3-4):205-12. doi: 10.1007/BF02795246.
Complex evaluation of secretin-pancreozymin test data with multivariate statistical methods was performed in patients with pancreatic insufficiency and in controls. The centroid, the nearest neighbour methods and the linear discriminant analysis led to the correct diagnosis in 84, 77 and 92%, respectively, when the overall responses to synthetic secretin plus cholecystokinin-octapeptide were evaluated. Results of individual stimulations were less precise. The linear discriminant analysis also successfully separated the patients with mild chronic pancreatitis from the healthy subjects, and seems to provide a useful support in medical diagnosis, particularly in controversial cases.
运用多变量统计方法对胰腺功能不全患者和对照组的促胰液素-胰酶泌素试验数据进行了综合评估。当评估对合成促胰液素加胆囊收缩素八肽的总体反应时，质心、最近邻法和线性判别分析分别在84%、77%和92%的情况下得出了正确诊断。个体刺激的结果准确性较低。线性判别分析还成功地将轻度慢性胰腺炎患者与健康受试者区分开来，并且似乎在医学诊断中提供了有用的支持，特别是在存在争议的病例中。
