Spatial Tracking Across Time (STAT): Tracking Neurons Across In-Vivo Imaging Sessions through Optimizing Local Neighborhood Motion Consistency

Abstract

Chronic calcium imaging has become a powerful and indispensable tool for analyzing the long-term stability and plasticity of neuronal activity. One crucial step of the data processing pipeline is to register individual neurons across imaging sessions, which usually extend over a few days or even months, and show various levels of spatial deformation of the imaged field of view (FOV). Previous solutions align FOVs of all sessions first and then register the same neurons according to their shapes and locations. However, the FOV registration is computational intensive, especially in the case of nonrigid case. Here we propose a cell tracking method that does not require FOV image registration. Specifically, the algorithm STAT (short for Stay Together, Align Together, and for Spatial Tracking Across Time) represents neurons from two sessions as two sets of neuronal centroids, uses point set registration (PSR) to find a spatially smooth transformation to align them while assigning correspondences. The optimization method iteratively updates between the general motion and individual neuron identity tracking, an idea seen in the computer vision literatures. Our method can be thought of as a specialization and simplification of these more general methods to calcium imaging neuron tracking. We validate STAT on datasets with simulated nonrigid motion that is hard to motion correct without extensive manual intervention. Next, we test STAT on experimental data from singing birds collected on three different days, and observe stable song-locked activity across days. An example use case of this package is reference [5].

Self-organization of songbird neural sequences during social isolation

“…We wondered if these birds, which learned a new syllable rapidly after tutoring, formed a de-novo HVC sequence for this new syllable, or perhaps used a pre-existing sequence. We were able to track neurons in our calcium imaging data throughout the course of tutoring ( Figure 3A , see Methods, Gu et al, 2023 ), enabling us to see what happens to neural activity during rapid changes in the song. We first extracted neural sequences associated with new post-tutoring syllables, then followed these neurons back in time to find that the sequence existed even prior to tutoring ( Figure 3B and C , see Methods).…”

“…Neuronal activity traces were extracted from raw fluorescence movies using a constrained non-negative matrix factorization algorithm, CNMF_E, that is specialized for microendoscope data by including a local background model to remove activity from out-of-focus cells ( Zhou et al, 2018 ). Custom software (Shijie Gu, Emily Mackevicius, Pengcheng Zhou) was used to extend the CNMF_E algorithm to combine batches of short files (BatchVer) and track individual neurons over the course of multiple days ( Gu et al, 2023 ).…”

“…We tracked the activity of populations of HVC neurons over multiple days using Spatial Tracking Across Time (STAT, Gu et al, 2023 ). This method builds off of previous methods ( Sheintuch et al, 2017 ), where individual cell pairs’ shape spatial correlation and distance are used to determine the correspondences between cells extracted from different sessions.…”