Detection of the area covered by neural stem cells in cultures using textural segmentation and morphological watershed

Abstract

Monitoring and evaluation of the dynamic of stem cells growth in culture is important in the regenerative medicine as a tool for cells population increasing to the size needed to therapeutic bprocedure. In this paper the automatic segmentation method of cells images from bright field microscope is proposed. It is based on the textural segmentation and morphological watershed. Textural segmentation aims at detecting within the image regions with intensive textural features, which refer to cells. Texture features are detected using local mean absolute deviation measure. Final, precise segmentation is achieved by means of morphological watershed on the gradient image modified by the imposition of minima derived from the result of rough segmentation. The proposed scheme can be applied to segment other images containing object characterized by their texture located on the uniform background.