2014/07/16 by Fabrizio De Vico Fallani, Martina Corazzol, Jenna R. Sternberg +3 · 26 citations
Biochemistry, Genetics and Molecular Biology · Computer Science · Medicine · Neuroscience · #Anatomy #Biological neural network #Biology #Calcium #Calcium imaging #Computational Drug Discovery Methods #Computer science #Embryonic stem cell #Functional Brain Connectivity Studies #Genetics #Granger causality #Internal medicine #Machine learning #Medicine #Neural dynamics and brain function #Neuroscience #Population #Sensory system #Spinal cord #Zebrafish #q-bio.NC
paper · pdf · doi:10.1109/tnsre.2014.2341632
published in IEEE Transactions on Neural Systems and Rehabilitation Engineering 23(3), 333-341 (Institute of Electrical and Electronics Engineers)
arxiv created 2014/07/16 · openalex publication_date 2014/08/06 · arxiv updated 2014/09/10 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05
The recent development of genetically encoded calcium indicators enables monitoring in vivo the activity of neuronal populations. Most analysis of these calcium transients relies on linear regression analysis based on the sensory stimulus applied or the behavior observed. To estimate the basic properties of the functional neural circuitry, we propose a network approach to calcium imaging recorded at single cell resolution. Differently from previous analysis based on cross-correlation, we used Granger-causality estimates to infer information propagation between the activities of different neurons. The resulting functional network was then modeled as a directed graph and characterized in terms of connectivity and node centralities. We applied our approach to calcium transients recorded at low frequency (4 Hz) in ventral neurons of the zebrafish spinal cord at the embryonic stage when spontaneous coiling of the tail occurs. Our analysis on population calcium imaging data revealed a strong ipsilateral connectivity and a characteristic hierarchical organization of the network hubs that supported established propagation of activity from rostral to caudal spinal cord. Our method could be used for detecting functional defects in neuronal circuitry during development and pathological conditions.