Automatic Segmenting and Classifying the Neural Stem Cells in Adherent Culturing Condition

The neural stem cells (NSCs) have a wide range of perspectives in clinical applications for neurology disorders due to their multi-potent potentials of differentiation. Automatic segment and classify the NSCs can be useful tools for biologist to monitor the progress of differentiation. In this paper, a hybrid image segmentation framework based on self-organizing map and watershed algorithm was applied to segment the NSCs in adherent culturing conditions. The cells shapes were analyzed using Fourier descriptors and classified using a feed-forward neural network. The results indicated that different shapes of NSCs in adherent culturing condition can be successfully segmented and classified based on these methods.