Abstract:Cellular shape dynamics are critical for understanding cell fate determination and organogenesis during development. However, fluorescence live-cell images of cell membranes frequently suffer from a low signal-to-noise ratio, especially during long-duration imaging with high spatiotemporal resolutions. This is caused by phototoxicity and photobleaching, which limit laser power and hinder effective time-lapse cell shape reconstruction, particularly in rapidly developing embryos. Here, we devised a new computational framework, EmbSAM, that incorporates a deep-learning-based cell boundary localization algorithm and the Segment Anything Model. EmbSAM enables accurate and robust three-dimensional (3D) cell membrane segmentation for roundworm Caenorhabditis elegans embryos imaged every 10 seconds. The cell shape data prior to gastrulation quantitatively characterizes a series of cell-division-coupled morphodynamics associated with cell position, cell identity, lineage, and fate, and can be accessed locally and online. The framework also exhibits potential in segmenting and quantifying the fluorescence labeling various cell-membraned-attached molecules in both wild-type and RNAi-treated embryos.

What problem does this paper attempt to address?

The paper attempts to address the issue of low signal-to-noise ratio (SNR) in fluorescence live-cell imaging of developing embryos, caused by phototoxicity and photobleaching, especially in scenarios requiring high spatiotemporal resolution for long-term imaging. Specifically, the paper proposes a new computational framework—EmbSAM (Embryo Segment Anything Model), which combines a deep learning-based cell boundary localization algorithm with the Segment Anything Model (SAM). EmbSAM enables accurate and robust 3D cell membrane segmentation under rapid imaging conditions (every 10 seconds) of Caenorhabditis elegans embryos. Additionally, this framework can quantitatively characterize the morphodynamic changes related to cell division during early embryonic development and demonstrates the potential to segment and quantify various fluorescently labeled cell membrane molecules in embryos with different genetic backgrounds. In summary, EmbSAM aims to improve the accuracy of cell boundary detection and segmentation under low SNR image conditions, thereby better studying the changes in cell behavior during development.

EmbSAM: Cell boundary localization and Segment Anything Model for fast images of developing embryos