Our Technology

Leveraging the newest advances in AI, our Human Foundation Model is trained on millions of unlabeled images for a ready-to-use, broadly applicable way to characterize cell morphology.

Extracting Information from Cell Images Two Ways

AI / deep learning derived features

Quantitative descriptions of cell features derived from deep learning models based on neural networks

Morphometric features

Quantitative analysis of cell and biological features, such as position, size, blebbing, and pigmentation

High-Dimensional Data You Can Interact With

Deepcell Software Suite

Morphometric and AI/deep learning derived features are combined and projected onto morphology UMAPs for easy visualization, cell group comparisons, and exploration.

How it works

Microfluidics

Dissociated single cell suspension is loaded into the instrument

How it works

High-res Single Cell Imaging

Label-free brightfield images are taken of each cell in flow

How it works

Real-time AI Classification

Human Foundation Model characterizes cells in real time and and provides additional cell attributes

How it works

Cell Sorting

Collect morphologically similar cell populations of interest with 6-way sorting

Technology Preprint

Realtime morphological characterization and sorting of unlabeled viable cells using deep learning

Deepcell AI

Self-supervised deep learning models
capture multi-dimensional features for cell morphology analysis from brightfield images

Deepcell Software Suite

Explore 3 data sets in the Deepcell Software Suite demo

Interested?

We’d love to explore how Deepcell’s technology can support your research.

Get in touch