A rapid segmentation method of cell boundary for developing embryos using machine learning with a personal computer

Abstract

In multicellular organisms, groups of cells cooperatively behave and change morphology, resulting in dynamic tissue deformation.In vertebrate gastrula, for example, coordinated migration of mesodermal cells occurs, including convergent extension, to establish complex structures inside the embryo to complete gastrulation (Keller & Danilchik, 1988). For understanding tissue deformation, it is necessary to describe the overall cell arrangement. Several methods for describing cell arrangement have been reported (Blanchard et al., 2009;Guirao et al., 2015;Merkel & Manning, 2017) that use strategies such as the relative position of the cell center (Guirao et al., 2015), the vertical cell shape by fitting of an ellipse (Blanchard et al., 2009), or direct tracing of the cell shape, including the apical cell area. These cell shape data are crucial for understanding cell movements, morphology, cell-cell interactions, and competitions during differentiation and migration (Lecuit & Le Goff, 2007;Oates et al., 2009). These cell shape data are usually obtained by labeling the cell membrane with either membrane-bound GFP in living cells or phalloidin staining in fixed cells. Then the fluorescent image is transformed into a line drawing for subsequent analyses. The critical points for segmentation methods are accuracy and time. The

In vivo imaging and regression analysis reveal cell-non-autonomous functions of Shroom3 during neural tube closure

“…Raw 3D images were projected to 2D via maximum intensity and underwent initial segmentation of cell boundaries using the FIJI plugin Tissue Analyzer (Aigouy et al, 2010; Aigouy et al, 2016). The segmentation of an initial frame was hand-corrected, and this hand-corrected segmentation was used to train a classifier using the programs CSML and EPySEG (Aigouy et al, 2020; Ota et al, 2021). CSML and EPySEG were used to generate segmentation for subsequent frames, which were then further hand-corrected in Tissue Analyzer.…”

Image analysis workflows to reveal the spatial organization of cell nuclei and chromosomes

“…In addition, these image analysis workflows are expected to inspire cell biologists beyond the study of the nucleus and its constituents. For instance, transposed at the cellular scale, workflow 1 or 4 could be applied to analyze the spatial distribution of vesicles or cytoplasmic bodies within a cell, using cell segmentation modules to create the initial surface object (see, for instance, but not exhaustive, references [66][67][68]).…”