1. Overview

The ASTEC suite is aimed at segmenting and tracking cells from series of membrane-labeled images acquired by the multiview SPIM (MuViSPIM) light-sheet microscope. Hence, several steps should be done sequentially. Astec philosophy is to segment one image and to forward propagate the segmentation in the series while identifying cell divisions.

  1. astec_fusion aims at fusing the 4 images per timepoint acquired by the MuViSPIM microscope.

    Tuning the microscope acquisition parameters (if not known) is described in Section 5.1.

    The imaged sample may move (with a large motion that can not be recovered by the registration method used here), either between the two points of view of one timepoint acquisition, or between two timepoints. astec_drift has been designed to recover these last transformations that will be used by ref:cli-fusion afterwards for fusion. Section 5.3 exemplifies this two-step fusion procedure.

  2. astec_mars aims at segmenting one timepoint (one fused image). It is basically a seeded watershed algorithm with the regional minima as seeds. Since some segmentation errors may occur, astec_manualcorrection provides some means to perform corrections (by fusing oversegmentations, and by specifying additionnal seeds).

  3. astec_astec aims at propagating a segmentation from one timepoint to the next, while recognizing cell divisions and building a cell lineage.

  4. astec_postcorrection aims at doing a first (automated) round of corrections of the propagated segmentation.

  5. astec_manualcorrection can be used as well to correct the propagated segmentation if required. Morphonet is also an alternative for that purpose.