Last year, 2014, we saw the firing of neurones in the brain of a zebrafish. That was a spectacular sight! You could see the dynamic firing of neurones extending over the various areas of the brain as the fish was "experiencing" or "doing" something.
However, that provided a macro view of brain areas affected, similar images, although not providing the same dynamics, can be obtained using PET and looking at the rate of sugar adsorption (which corresponds to the level of energy utilisation by neurones).
We know that the brain does not split into drawers each of them in charge for a specific task. There is some prevalence, like saying that in human brain the occipital part of the cortex is involved in processing vision. However, that same part gets involved in hundreds of other activities.
To pinpoint the neurones and the sequence of their involvement one has to film at a much higher speed. This is what researchers at Duke university have been able to do achieving a time resolution of the order of 0.2 ms. That is fast enough to result in a real time monitoring of the firing of neurones in animals (the basics are the same, whether it is a fruit fly neurone or one of Einstein).
To do this the team at Duke has encoded in neurones, using genetic engineering, voltage indicators that can sense individual action potentials in live animal cells and a protein that can amplify the signal of those indicators through fluorescence. It is this fluorescence that is captured and filmed.
Basically, they have managed to insert a tiny lamp in each neurone and to power this lamp with the same voltage potential used by the neurone to communicate with other neurones.