Figure 2

Image Processing Algorithm for Quantification of Neuronal Morphology. a) Flow diagram that illustrates the extraction of key morphological features from single colour images of neurons stained with antibodies to either the neuron-specific marker betaIII-tubulin or the dendrite-specific marker MAP2. Numbers refer to steps in the algorithm description (Main Text: Implementation). A summary of the analysis algorithm is provided in the Additional File 1. b) Example of feature extraction from cultured primary hippocampal neurons. The majority of neuronal structures are identified accurately by this procedure. The analysis algorithm quantifies total neurite length by counting pixels of the skeletonized neurites (green) and total neuronal cell body area per field of view (red). In addition, the algorithm identifies and counts the majority of cell bodies, neurite endpoints (cyan) and neurite-cell body attachment points (yellow), as well as the total staining intensities per field. These values are used to deduce additional measurements (see text for details).