CalTrig: A GUI-Based Machine Learning Approach for Decoding Neuronal Calcium Transients in Freely Moving Rodents

Area of Science:

• Neuroscience
• Computational Biology
• Data Science

Background:

• Advances in in vivo calcium (Ca²⁺) imaging allow single-neuron studies in freely moving animals.
• Tools like Minian and CalmAn generate large numerical datasets (CalV2N) from Ca²⁺ signals.
• Analyzing CalV2N data presents challenges in data integration, quality assessment, and transient identification.

Purpose of the Study:

• Introduce CalTrig, an open-source graphical user interface (GUI) for post-CalV2N data analysis.
• Address challenges in integrating data streams, evaluating output quality, and identifying Ca²⁺ transients.
• Provide a user-friendly tool for neuroscientists without programming expertise.

Main Methods:

• Developed CalTrig, a GUI tool integrating Ca²⁺ imaging, neuronal footprints, Ca²⁺ traces, and behavioral tracking.
• Implemented synchronized visualization and efficient Ca²⁺ transient identification.
• Evaluated four machine learning models (GRU, LSTM, Transformer, Local Transformer) for Ca²⁺ transient detection.

Main Results:

• The GRU model demonstrated the highest predictability and computational efficiency for Ca²⁺ transient detection.
• GRU model achieved stable performance across different training sessions, animals, and brain regions.
• CalTrig successfully integrates multiple data streams and facilitates quality evaluation of CalV2N outputs.

Conclusions:

• CalTrig offers a flexible and accurate solution for Ca²⁺ transient identification through integrated manual, parameter-based, and machine learning methods.
• The tool's user-friendly interface and low computational demands make advanced data analysis accessible.
• CalTrig facilitates deeper exploration of brain function, hypothesis generation, and understanding of neurological disorders.