Skip to main content
. 2010 Oct 13;10:115. doi: 10.1186/1471-230X-10-115

Figure 6.

Figure 6

Decision tree for the differentiation of various stages fibrosis using four serological markers. Briefly, in a training group of 73 samples, samples were first divided into three groups (normal & milder fibrosis, milder & significant fibrosis, significant fibrosis & early cirrhosis) by marker "PT" (cutoff value = 12.00, 13.60 and 14.70, respectively); The normal & milder fibrosis group was then correctly classified by marker "Pre" (cutoff value = 0.23); the milder & significant fibrosis group was then correctly classified by marker "Prx II" (cutoff value = 0.80); the significant fibrosis & early cirrhosis group was then classified by marker "HA" (cutoff value = 381.74) and all the early cirrhosis samples (9/9) were correctly classified. The correct prediction percentage of the algorithm for normal control, milder fibrosis, significant fibrosis and early cirrhosis was 100%, 88.9%, 95.2% and 100%, respectively.