Statistical modeling of neural spike activity is a challenging problem because accurate measurements are not easily accessible as they depend on background noise and on other neighboring neuronal activity. Recent technological advancements in fluorescence microscopy, however, make possible to examine thoroughly neuronal responses to external stimuli through calcium imaging data. Such data consists of intracellular fluorescent calcium traces recorded over a long period of time for a certain area of the brain. In this work, after a preliminary exploratory data analysis, we develop time series methodology for identifying spiking activity of several neurons at different brain depths. We also exploit clustering techniques to group neurons that exhibit similarities in the activation patterns.