Robust tuning of Expanding Active Contour Model for tracking lower limb vein in ultrasonographic image

摘要

This paper reports on tuning the performance of an automated measurement system that we integrated for vein tracking in ultrasonographic image. We target to monitor the venous flow in B-mode ultrasonograhy and set our technical challenge to assemble image processing algorithms, including the Expanding Active Contour Model, to extract the vein contour. Parameter Design method of Robust Engineering is applied to analyze and select factor's optimal level. Confirmation test was conducted to compare the optimal and initial conditions. The results show that the selected level setting increases the tracking ability, yet, consideration is required on the repeatability of the measurement system.