The Establishment And Performance Verification Of The Model Of Liver Fibrosis With Non-Invasive Diagnosis

Objective:To establish suitable for non-invasive diagnosis of liver fibrosis patients in this region model, and verify the diagnosis efficiency of the model.Methods:Collected from 2011 to 2014 in Guangdong hospital of hospitalization and cases of liver biopsy examination, serologic test results of retrieval cases at the same time, including HA, PIIINP, LN, CIV, ALT, AST, GGT, TBIL, DBIL, propagated, ADA, UREA, ALP, APOA1, PT, PLT.354 cases with collecting data complete cases. Of serological indexes and binary logistic regression analysis between pathological diagnosis, diagnosis model are obtained. The complete cases verified model diagnostic performance data collected.Results:Through regression model analysis, it is concluded that the diagnosis model of forecasting formula is:In (p1/1-p1)=12.445+0.061×LN-0.011×CIV-0.313×ALB+0.313×ADA-0.011×PLTIn (p2/1-p2)=6.443+0.098×PIIINP-0.002×CIV+0.014×GGT-0.161×ALB +0.148×ADA-0.01×PLTIn (p3/1-p3)=-0.215+0.022×HA+0.113×PIIINP-0.004×AST-0.011×ALB +0.015×ALP-0.017×PLTIn (p4/1-p4)=0.394+0.011×HA+0.002×CIV-0.012×ALB+0.013×ALP-0.02×PLTConelusion:Through the establishment of the diagnosis model can effectively reduce the liver fibrosis patients puncture in the diagnosis of probability.