Identification of Neuronal Ensembles from Similarity Maps Using Singular Value Decomposition

Abstract

Abstract Neuronal ensembles, defined as groups of neurons with coordinated activity, have been proposed as the basic motifs for brain computations such as perception, memory, and movement. This chapter delves into the implementation of Singular Value Decomposition (SVD) applied to similarity maps of neuronal population vectors as an analytical method to identify neuronal ensembles from calcium imaging recordings. At its core, SVD is a mathematical tool that allows the factorization of data matrices. In the context of neuronal ensemble identification, SVD could be applied to similarity maps constructed from multidimensional population vectors, where the dimensionality of the array is defined by the total number of observed neurons; thus similar population vectors are extracted as different SVD factors. Population vectors depict neuronal activity states as screenshots of active and inactive neurons at different time windows. The goal of this chapter is to provide the conceptual intuition for the implementation of SVD from similarity maps to identify neuronal ensembles. The identification of neuronal ensembles could shed light to understand the role of neuronal population dynamics in brain function.