Fast image analysis tools for digital signal processing of single-cell data streams.
110
CoyleLab-UW-Madison/cellstream:
cellstreamis a PyTorch-accelerated Python image processing package that provides a suite of tools for single-cell analysis of frequency-domain and time-frequency domain features in fluorescence microscopy data. Initially designed for use with programmable reaction diffusion systems and genetically-encoded oscillator circuits (GEOs), the tools can also be applied to a wide range of dynamic cellular systems. Continuous wavelet transforms (CWT) make use of the excellent ssqueezepy package. GPU functionality is available but not required.
This docker image is suitable for GPU-accelerated processing of microscopy data. It includes a preconfigured python environment with cellstream, pytorch 2.7.0, cuda 11.8, and torch-scatter configured. It also includes cellpose-SAM for generating masks.
The image can be run locally, taking advantage of a cuda-compatible GPU on a linux or windows machine. Alternatively, it can be run inside a remote cluster, such as one managed by HTCondor.
Content type
Image
Digest
sha256:3f813491e…
Size
3.7 GB
Last updated
about 1 year ago
docker pull eweix/cellstream:0.1.0-headless