Cellpose | Python, GUI, Deep Learning | GitHubDoc | - Overview: A generalist deep learning algorithm for cell segmentation across diverse cell shapes, sizes, and cellular modalities.
- How to Bridge: Performs single-cell or nuclear segmentation on multi-stained tissue images via its GUI or Python API, then exports standard cell masks in TIFF or GeoJSON format to integrate into STOmics analysis tools for high-precision cell binning.
| cellpose.svg |
StarDist | Python, Java, Deep Learning | GitHubDoc | - Overview: A star-convex object detection model optimized for dense, star-convex shapes like cell nuclei.
- How to Bridge: Ideal for nuclear-stained (e.g., DAPI) raw tissue images. Output nuclear masks exported in TIFF or GeoJSON format can be imported into STOmics analysis tools to assist in nuclear-guided cell binning.
| stardist.svg |
Omnipose | Python, Deep Learning | GitHubDoc | - Overview: A deep learning-based segmentation algorithm designed for extremely varied cell shapes, bacterial microscopy, and dense tissues.
- How to Bridge: Suitable for non-standard or highly complex cell morphology datasets where generalist models like Cellpose fall short. Output nuclear masks can be directly imported into STOmics analysis tools.
| omnipose.svg |
DeepCell | Python, CLI, Deep Learning | GitHubDoc | - Overview: A deep learning library featuring pre-trained models (like Mesmer) for cell segmentation in tissue imaging.
- How to Bridge: Provides robust segmentation across diverse tissue modalities. Output nuclear or membrane masks can be converted to standard cell segmentation formats for STOmics tools ingestion.
| deepcell.svg |
