CaliAli, a tool for long-term tracking of neuronal population dynamics in calcium imaging

Abstract

AbstractNeuron-tracking algorithms exhibit suboptimal performance in calcium imaging when the same neurons are not consistently detected, as unmatched features hinder intersession alignment. CaliAli addresses this issue by employing an alignment-before-extraction strategy that incorporates vasculature information to improve the detectability of weak signals and maximize the number of trackable neurons. By excelling in neural remapping and high spatial overlap scenarios, CaliAli paves the way toward further understanding long-term neural network dynamics.