The olfactory bulb (OB) is the first stage of olfactory information processing in the brain. On its way to the cortex, odor information is encoded in spatiotemporal maps of activated loci in the bulb, and these are known as input maps. Using optical recording techniques, experimental neuro-scientists can reveal the input maps by applying response-dependent fluorescent dyes to the OB of rodents and imaging the odorant-evoked responses. A hard signal analysis task then emerges (i.e., to handle the resulting sequence of spatial patterns). We suggest that the intricate spatiotemporal patterning of event-related dynamics can be described fully by using a manifold learning approach that merges both the spatial and temporal aspects of the response into a single visualization scheme.