lacss 0.13.1

Tools for cell segmentation

LACSS

   pip install lacss

LACSS is a deep-learning model for single-cell segmentation from microscopy images. See references below:

• IEEE TMI (2023) doi:10.1109/TMI.2023.3312988
• Communications Biology 6,232 (2023)

What's new (0.11)

You can now deploy the LACSS predictor as an GRPC server:

   python -m lacss.deploy.remote_server --modelpath=<model_file_path>

For a GUI client see the Trackmate-Lacss project, which provides a FIJI/ImageJ plugin to perform cell segmentation/tracking in an interactive manner.

Why LACSS?

LACSS is designed to utilize point labels for model training. You have three options: (1) Label each cell with a single point, (2) label each cell with a single point and then label each image with a binary mask that covers all cells, or (3) Label each cell with a separate segmentation mask (as in standard supervised training). You can of course also combined these labels in any way you want.

Method	Data(left) / Label(right)
Point
Point + Mask
Segmentation

How to generate point label?

If your data include nuclei counter-stain, the easist way to generate point label for your image is to use a blob detection algorithm on the nuclei images:

Give It A Try:

Model Training

• Supervised Training
• With point label + mask
• With point label only

Model Inference

• Inference

Documentation

API documentation