Figure 4
From: A fast and robust hepatocyte quantification algorithm including vein processing
ROC-like Curve for Hough Transformation Threshold. ROC-like curve has sensitivity as x axis and false positive fraction (FPF) as y axis. ROC-like curves are built for images D 4-CV 1 and D 8-CV 1 for threshold values [0.1, ..., 0.9] with step = 0.1. The optimal values of the relevance threshold for Hough transformation lie within interval [0.5, 0.6], which corresponds sensitivity ≥ 90% and FPF ≤ 10%.