Cell Segmentation with Adaptive Region Growing

The cell segmentation presented in this paper is part of current research work on an automated cell tracking project. Its aim is the tracking of the movement of representative cells in a 3D collagen gel in order to determine the activity of the cell after the injection of certain medical substances. A major requirement for this project is an efficient and robust method for the segmentation of cell images. Here we present an automated method based on region growing with a local adaptive threshold. For preprocessing we use an optimized rational filter for edge preserving noise smoothing. The experimental results show that the presented method is suitable for the segmentation of melanoma cells (tumor cells) as well as lymphocytes (blood cells) in a 3D collagen gel.