Segmentation and analysis of round and linear F-actin assemblies. (A) Example of image analysis pipeline: inverted contrast image of mNGMA-labeled presynaptic actin in live dissected larvae. A two-class (round and linear F-actin) WEKA classifier was applied. (B) Product of the WEKA-based segmentation of linear actin is a 32-bit probability map image, which is converted to 16-bit image to make a binary mask (8-bit). The binary mask was further skeletonized, and circularity (0–0.2) and size (0.02 µm²–infinity) cutoffs were implemented at the particle analysis step to distinguish bona fide linear F-actin. (C) 32-bit probability map image of round F-actin was similarly converted to 16-bit image, which was used to make a binary mask of the structures. The binary mask was used in particle analysis of structures with a size higher than 0.0018 µm², i.e., 1 pixel². See additional description in Materials and methods.