Segmentation of medical image sequence by parallel active contour

Abstract

This paper presents an original approach for detecting and tracking of objects in medical image sequence. We propose a multi-agent system (MAS) based on NetLogo platform for implementing parametric contour active model or snake. In NetLogo, mobile agents (turtles) move over a grid of stationary agents (patches). In our proposed MAS, each mobile agent represents a point of snake (snaxel) and it minimizes, in parallel with other turtles, the energy functional attached to its snaxel. Then, these turtles move over the image represented by a grid of patches. The set of these agents is supervised by Observer, the NetLogo global agent. In addition, Observer loads successively the frames of sequence and initializes the turtles in the first frame. The efficiency of our system is shown through some experimental results. © 2011 Springer Science+Business Media, LLC.