Multi-day neuron tracking in high-density electrophysiology recordings using earth mover's distance.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Electrophysiology

Background:

• Tracking individual neurons over time is essential for understanding neural plasticity, learning, and adaptation.
• High-density electrophysiology probes like Neuropixels offer potential for long-term neuronal recordings.
• Existing neuron tracking methods often rely on firing statistics, which can be unreliable due to neuronal signal loss and tissue drift.

Purpose of the Study:

• To develop a novel neuron tracking method independent of firing statistics.
• To enable accurate identification of the same neurons across multiple days despite tissue movement.
• To improve the reliability of longitudinal studies in neuroscience.

Main Methods:

• Proposed a method based on between-day non-rigid alignment of spike-sorted clusters.
• Utilized visual receptive field measurements to verify same-cell identity in mice.
• Applied the method to datasets with recordings separated by 1 to 47 days.

Main Results:

• The proposed method successfully identified the same neurons across multiple days.
• Achieved an average neuron recovery rate of 84%.
• Demonstrated robustness across recording intervals ranging from 1 to 47 days.

Conclusions:

• The developed neuron tracking method offers a reliable approach for longitudinal neuronal analysis.
• This technique overcomes limitations of existing methods by not relying on firing statistics.
• Enables more accurate studies of neuronal changes during learning and adaptation.