CascadeTorch: Calibrated spike inference from calcium imaging data using PyTorch

Cascade translates calcium imaging ΔF/F traces into spiking probabilities or discrete spikes.

Cascade is described in detail in the main paper. There are follow-up papers which describe the application of Cascade to spinal cord data and the application of Cascade to GCaMP8.

Cascade's toolbox consists of

• A large and continuously updated ground truth database spanning brain regions, calcium indicators, species
• A deep network that is trained to predict spike rates from calcium data
• Procedures to resample the training ground truth such that noise levels and frame rates of calcium recordings are matched
• A large set of pre-trained deep networks for various conditions (additional models upon request)
• Tools to quantify the out-of-dataset generalization for a given model and noise level
• A tool to transform inferred spike rates into discrete spikes

Get started quickly with the following Colaboratory Notebook:

• Spike inference from calcium data (Colaboratory Notebook)
• Upload your calcium data, use Cascade to process the data, download the inferred spike rates.
• Spike inference with Cascade improves the temporal resolution, denoises the recording and provides an absolute spike rate estimate.
• No parameter tuning, no installation required.
• You will get started within few minutes.

Getting started

Without installation

If you want to try out the algorithm, just open this online Colaboratory Notebook. With the Notebook, you can apply the algorithm to existing test datasets, or you can apply pre-trained models to your own data. No installation will be required since the entire algorithm runs in the cloud (Colaboratory Notebook hosted by Google servers; a Google account is required). The entire Notebook is designed to be used by researchers with little background in Python, but it is also the best starting point for experienced programmers. Try it out - within a couple of minutes, you can start using the algorithm!

With a local installation (Ubuntu/Windows)

If you want to modify the code, integrate the algorithm into your existing pipeline (e.g., with CaImAn or Suite2P), or train your own networks, you will need a local installation.

Although Cascade is based on deep networks, GPU support is not required. Model training runs smoothly on CPUs (although GPUs can speed up the process). Therefore, installation is much simpler than for typical deep learning toolboxes that depend on GPU-based processing.

Inference has been tested successfully with Torch versions between 2.4 and 2.9 on Colab, Ubuntu, and Windows. See setup.py for a full list of requirements, or navigate to the CascadeTorch folder in your environment and run: pip install .

Feedback about problems with configurations and operating systems (also positive feedback about working environments) is welcome. Please submit issues, e-mails, or pull requests.

Updates, FAQs, further info:

Check the parent CASCADE repository. FAQs and updates are updated only there for simplicity.

References

Please cite as primary reference for Cascade:

Rupprecht P, Carta S, Hoffmann A, Echizen M, Blot A, Kwan AC, Dan Y, Hofer SB, Kitamura K, Helmchen F*, Friedrich RW*, A database and deep learning toolbox for noise-optimized, generalized spike inference from calcium imaging, Nature Neuroscience (2021). (* = co-senior authors)

And the following papers specific for models trained with GCaMP8 and spinal cord data, respectively:

Rupprecht P, Rózsa M, Fang X, Svoboda K, Helmchen F. Spike inference from calcium imaging data acquired with GCaMP8 indicators, bioRxiv (2025).

Rupprecht P, Fan W, Sullivan S, Helmchen F, Sdrulla A. Spike rate inference from mouse spinal cord calcium imaging data, J Neuroscience (2025).