Tracking the Same Neurons across Multiple Days in Ca²⁺ Imaging Data

Area of Science:

• Neuroscience
• Computational Biology
• Data Science

Background:

• Calcium (Ca²⁺) imaging allows recording neuronal activity over weeks.
• Longitudinal analysis of neural codes requires accurate cell registration across sessions.
• Current methods lack quantitative accuracy evaluation, risking data misinterpretation.

Purpose of the Study:

• To develop an automated method for accurate cell registration across multiple calcium imaging sessions.
• To provide quantitative confidence estimates for registered cells.
• To enable reliable longitudinal analysis of neuronal populations.

Main Methods:

• Developed a probabilistic method for automatic cell registration.
• Applied the method to large-scale Ca²⁺ imaging data from mouse hippocampus and cortex.
• Validated registration accuracy and scalability over multiple sessions.

Main Results:

• The probabilistic method achieved higher accuracy than existing routines.
• Estimated error rates for cell registration were below 5%.
• The method demonstrated scalability for numerous sessions and cells.

Conclusions:

• The developed method enables reliable longitudinal analysis of individual neurons over extended periods.
• Accurate cell registration is crucial for understanding neural codes.
• This approach advances the analysis of large-scale neural dynamics.