| Literature DB >> 29782491 |
Philipp Berens1,2,3, Jeremy Freeman4,5, Thomas Deneux6, Nikolay Chenkov7, Thomas McColgan7, Artur Speiser8, Jakob H Macke8,9, Srinivas C Turaga5, Patrick Mineault10, Peter Rupprecht11,12, Stephan Gerhard11, Rainer W Friedrich11,12, Johannes Friedrich13, Liam Paninski13, Marius Pachitariu5,14, Kenneth D Harris14, Ben Bolte15, Timothy A Machado13, Dario Ringach16, Jasmine Stone5,17, Luke E Rogerson1,2,3, Nicolas J Sofroniew5, Jacob Reimer18,19, Emmanouil Froudarakis18,19, Thomas Euler1,2,3, Miroslav Román Rosón1,2,20, Lucas Theis21, Andreas S Tolias18,19,22, Matthias Bethge2,3,19,23.
Abstract
In recent years, two-photon calcium imaging has become a standard tool to probe the function of neural circuits and to study computations in neuronal populations. However, the acquired signal is only an indirect measurement of neural activity due to the comparatively slow dynamics of fluorescent calcium indicators. Different algorithms for estimating spike rates from noisy calcium measurements have been proposed in the past, but it is an open question how far performance can be improved. Here, we report the results of the spikefinder challenge, launched to catalyze the development of new spike rate inference algorithms through crowd-sourcing. We present ten of the submitted algorithms which show improved performance compared to previously evaluated methods. Interestingly, the top-performing algorithms are based on a wide range of principles from deep neural networks to generative models, yet provide highly correlated estimates of the neural activity. The competition shows that benchmark challenges can drive algorithmic developments in neuroscience.Entities:
Mesh：
Substances：
Year: 2018 PMID： 29782491 PMCID： PMC5997358 DOI： 10.1371/journal.pcbi.1006157
Source DB: PubMed Journal: PLoS Comput Biol ISSN： 1553-734X Impact factor: 4.475