Getting started#

Five short tutorials that take you from an unopened microscopy file to a growth-rate curve. They are meant to be read in order — each one builds on the last.

Every tutorial is a real Jupyter notebook. Click the Colab badge at the top of any of them to run it in your browser with nothing installed, or find them under docs/tutorials/ in the repository.

1. Open your first file

One function for ND2, CZI, TIFF stacks and folders of TIFFs. Read metadata without loading pixels, and pull out a position as a time series.

1. Open your first file
2. The sequence model

The (T, H, W, C) convention and numpy-style slicing — subsample, crop and pick channels without copying a single pixel.

2. The sequence model
3. Look at your data

The interactive Jupyter viewer, contact sheets, scale bars, timestamps, and exporting an annotated video.

3. Look at your data
4. Calibration and units

Declare pixel size and frame interval once, and get results in µm² and hours instead of px² and frames.

4. Calibration and units
5. Segment and quantify

The payoff: deep-learning segmentation, measurement, artefact filtering and a population growth rate — on real data, GPU optional.

5. Segment and quantify

Before you start, make sure acia is installed — see Installation. The notebooks install it themselves if you run them on Colab.