Serving algorithms

Serving algorithms#

Any Imaging Server Kit algorithm can be served as a web API using a built-in FastAPI server. This turns algorithms into web servers that you can interact with from Napari, QuPath, or Python via HTTP requests.

Serving an algorithm is particularly useful when:

  • You want to run computations on a remote machine and visualize the results on your local machine.

  • You want to use algorithms in QuPath via the qupath-extension-serverkit extension.

Using sk.serve#

Let’s consider the threshold algorithm once again. You can serve this algorithm by passing it to sk.serve().

import imaging_server_kit as sk
import skimage.data

@sk.algorithm(
    name="Intensity threshold",
    parameters={"threshold": sk.Integer(name="Threshold", min=0, max=255, default=128)},
    samples=[{"image" : skimage.data.coins()}],
)
def threshold_algo(image, threshold):
    mask = image > threshold
    return sk.Mask(mask, name="Binary mask")

if __name__ == "__main__":
    sk.serve(threshold_algo)  # <- Serve the algorithm

Important

sk.serve() won’t work in a Jupyter notebook. You need to run the code as a Python script.

Calling sk.serve() starts a local FastAPI server exposing your algorithm via a set of predefined routes.

Server-running

By default, the server is hosted at http://localhost:8000. If you navigate to this URL in a browser, you will see the familiar algorithm doc page.

Once the server is running, you can connect to it from Napari, QuPath, or directly from Python.

Connecting from Napari#

The Napari Server Kit plugin provides a widget for connecting to algorithm servers.

To try it out, open a new terminal window, start Napari, and then open the plugin from Plugins > Server Kit (Napari Server Kit). In the Server URL field, enter: http://localhost:8000, and press Connect. You should see your threshold algorithm listed, with all related functionalities (load samples, run algorithm, access documentation) available.

Connecting from QuPath#

You can also connect to algorithm servers via QuPath through the qupath-extension-serverkit extension. The QuPath extension can handle compatible algorithm outputs, including segmentation masks, bounding boxes, vectors, points, and classification tasks.

Summary#

  • Use sk.serve() to expose any Imaging Server Kit algorithm as a FastAPI web service.

  • The server runs locally at http://localhost:8000 by default.

  • You can connect to algorithm servers from Napari or QuPath.

Next steps#

In the next section, we will explore how to interact with Server Kit algorithms directly from Python.

Contents