Elastic source selection for in vivo imaging of neuronal ensembles

摘要: Advances in microscopy and biochemistry now allow investigators to image the calcium dynamics of hundreds thousands neurons awake behaving animals. However, as speed resolution such techniques rapidly increase, so do dimension complexity data collected. ICA has been widely employed reveal independent non-Gaussian sources underlying large sets consisting mixed sources. We apply a recently developed sparse regression method, Elastic Net (ENET), columns mixing matrix Independent Component Analysis (ICA) procedure. This method automatically selects only those relevant dependent variable interest. Further, because is linear operator, we can easily project “relevance filtered” back into native space for interpretation. demonstrate utility this on 3D imaging collected from optic tectum an larval zebrafish watching prey-like stimulus.