CaImAn an open source tool for scalable calcium imaging data analysis.

Area of Science:

• Neuroscience
• Computational Biology
• Bioimaging

Background:

• Advances in fluorescence microscopy necessitate robust, automated, and scalable analysis pipelines for in-vivo brain monitoring.
• High data rates from modern imaging techniques require efficient pre-processing methods.

Purpose of the Study:

• To introduce CaImAn, an open-source library for calcium imaging data analysis.
• To provide automated and scalable solutions for common pre-processing challenges in neural data analysis.

Main Methods:

• CaImAn offers automatic motion correction, neural activity identification, and cross-session registration.
• The library is designed for scalability across various computing platforms, from laptops to HPC clusters.
• It supports both two-photon and one-photon imaging modalities and enables real-time analysis.

Main Results:

• CaImAn demonstrates minimal user intervention and high scalability.
• Performance benchmarking on nine mouse datasets showed near-human accuracy in detecting active neurons.
• The library is suitable for analyzing large datasets collected over extended periods.

Conclusions:

• CaImAn provides a reliable, automated, and scalable solution for calcium imaging data analysis.
• The tool significantly aids in pre-processing complex neural datasets.
• It represents a valuable resource for neuroscience research requiring efficient in-vivo brain activity monitoring.