CellSeg is a powerful cell segmenter powered by deep learning.

Deep Learning

CellSeg is trained on tens of thousands of different cell images.

Open Source

CellSeg is fully open-source and built to be easily extendable.

Cross Platform

CellSeg is supported on Windows, Mac, and Linux.

Flexible Implementation

CellSeg segments cells of all different tissue types from flourescent microscopy images. Nuclear stains are best, but H&E works too! Point to TIFF, PNG, or JPG images and segment on your machine today.

Powerful Built-in Features

CellSeg 1.0 comes with many essential features required by biologists. Dilating cell regions, quantifications, output to CSV and FCS, conversion to ImageJ ROIs, visualizations, and more! Add your own easily.

Frequently Asked Questions

How does CellSeg work?

CellSeg is trained with thousands of annotated cell images in different sizes, tissues, and lighting conditions. CellSeg has learned to be robust to many different cell types without setting parameters that need to be learned. CellSeg is open-source, so you can see the code on Github if you click See Source!

How can I cite CellSeg?

Please see How to cite section of the README.md on our GitHub repo.

What platforms and specifications do you support?

CellSeg 1.0 is supported on Windows, Mac, and Linux. Due to CellSeg’s implementation, it is recommended that the running computer has at least 32 GB of memory (RAM), especially for larger images / denser tissues. Future releases of CellSeg will be more memory efficient.