Skip to main content

Table 2 Comparing input and output variables in three sets

From: Identification of the risk for liver fibrosis on CHB patients using an artificial neural network based on routine and serum markers

Variables Patients in the training set, n = 226 Patients in the validation set, n = 113 Patients in the testing set, n = 116 χ2(p)
Age(years) 30(16-62) 35(16-75) 32(16-74) 1.741(0.419)
WBC(109/L) 5.4(1.3-18.1) 5.5(2.5-16.8) 5.4(3.0-11) 0.231(0.891)
RBC(1012/L) 4.7(2.6-4.0) 4.7(3.6-5.7) 4.6(3.7-5.6) 0.310(0.856)
Platelet(109/L) 165(53-246) 170(13-350) 185(46-345) 4.965(0.083)
ALT(U/L) 87(16-471) 80(11-536) 72(6-507) 2.475(0.290)
AST(U/L) 48(11-413) 42.5(15-375) 39(11-361) 1.264(0.531)
GGT(U/L) 29(6-421) 33(6-288) 34(10-473) 3.247(0.197)
ChE(U/L) 8402(50-20427) 7674(145-11374) 7258(360-12648) 3.051(0.217)
Alb(g/L) 44.8(10.9-47.0) 44.2(11.8-50.7) 44.1(31.3-52.6) 0.412(0.814)
TBIL(μmol/L) 14.1(3.6-216) 15.3(3-196) 13.2(3- 134) 3.650(0.161)
Fibrosis
F(0-1)/F(2-4)
166/60 83/30 80/36 0.657(0.720)