Setting up Napari

To train new models with CryoViT, you will need to provide a labeled training dataset.

For this, we recommend using napari, an open-source, Python-based medical image viewer with a rich plugin ecosystem.

Tip

We recommend installing napari in its own separate environment, as described in the napari installation instructions and below.

Then, you can install the CryoViT napari plugin using pip in the same environment, as an alternative to the Plugin Manager.

CryoViT provides a napari plugin to help create and manage training datasets, which can be installed directly from PyPI, or through the Plugin Manager within napari.

Note

If you do not plan to train your own models, and only want to use the provided pre-trained models, you do not need to install napari or the CryoViT napari plugin.

Installing Napari

First, create a new conda environment for napari:

$ conda create -n cryovit-napari
$ conda activate cryovit-napari

Then, install napari using mamba (recommended) or conda:

$ mamba install -c conda-forge napari
# or
$ conda install -c conda-forge napari

Launch napari by activating the environment and running the napari command:

$ conda activate cryovit-napari
$ napari

Finally, check out the napari documentation for help on using the application.

Installing the CryoViT Napari Plugin

Using the Plugin Manager

To install the CryoViT napari plugin using the Plugin Manager:

1. Launch napari

2. Open the Plugin Manager from the “Plugins” menu under “Install/Uninstall Plugins”.

3. Search for cryovit-napari and click “Install”

Using pip

You can install the CryoViT napari plugin using pip with the napari environment activated:

$ conda activate cryovit-napari
$ pip install cryovit-napari