Workflow of the image analysis pipeline. Image processing upstream of FMAj consists of three steps. 3D time-lapse image data acquired by laser scanning confocal microscopy (a) are concatenated using the TLM-Converter custom software (step 2). Another custom software, called TLM-2D-Explorer (step 3), generates MIPs of 3D stacks, each of which represents a time point. The workflow in FMAj consists of 5 major steps. Note in the text, steps 4 and 5 are referred to as module 1, 6 and 7 as module 2 and 8 as module 3. (step 4) The user interactively annotates the anatomical reference points like anterior-posterior axis, left-right axis and the temporal reference points, such as the time of head eversion (b). The annotation details and metadata collected from the microscopic images are stored in the database. (step 5) To facilitate visual comparison of different time-lapse datasets, the anterior-posterior axis is rotated to be shown from left to right in the horizontal orientation. (step 6) Manual and semi-automated segmentation methods based on level-set help to define the contours of muscle fibers recorded in the green channel. In an optional step, nuclei recorded in the red channel can be segmented using an automated method. (c) Magnification of the segmented muscle indicated by an arrowhead in (b) and its nuclei. The ROIs corresponding to muscle fibers and their nuclei are stored in the database. (step 7) Shape and size features are calculated from the boundaries of the ROIs and stored into the database. ROIs are manually assigned muscle types for subsequent tracking and time-series analysis. (step 8) Tracks of ROIs along with their feature values are retrieved from the database to perform a statistical analysis of muscle phenotypes.