Three years after Azevedo et al (2009) dissolved a couple of human brains into a soup in order to obtain a more accurate estimate of the total number of neurons in the human brain ( a still staggering figure of 86.1 +/- 8.1 billions), Dr. Sebastian Seung and his team in MIT decided to go the other way in order to figure out how exactly these 86 bn neurons are connected with each other.
It is generally accepted that the exact wiring of all these neurons, the
connectome, is that which really constraints neural activity - which neurons interact with which - and by extension that which underlies all information processing and computing taking place in the brain. And many researchers like Seung take this further and support that it is the connectome that really makes us who we are.
So far, many ways have been tried in order to map the brain connectome: from diffusion MRI studies, which provide a large scale picture of neuron connectivity - but with very little resolution, to multiple stains microscopy techniques, which can provide good resolution but which are, limited to just a few cells. Yet again, the estimated number of synapses of the human brain is unimaginable; just the neocortex (20 bn neurons) has about 150 trillions of them (Pakkenberg 1997; 2003) and this definitely makes the mapping of a single brain connectome a titanic endeavour.
For this reason Seung's team decided to focus their investigation on the mapping of a retina that was sliced by a microtome and scanned by electron microscopy in order to digitally reconstruct its neurons. However, this is a task that no computer can do alone (at least today) and it takes up to 50hrs in order to reconstruct a single neuron. Computers are prone to many errors when analysing such data. Hence, it would take centuries to reconstruct an adequate number of them if it weren't for the power that can be harnessed from the crowds!
Seung and his team devised the EyeWire platform, a platform which allows 1000s non-expert gamers to spend their time - literally - constructively by mapping the projections of certain retinal neurons called
JAM-B cells. These cells are characterised by direction-selectivity and hence respond specifically to upward motion (in the visual field) and can by distinguished from others by a protein called Junctional Adhesion Molecule B (JAM-B) which gave them their name. Interestingly the cells themselves show some topographic asymmetry and this raises many questions about the relationship between their structure/connections and their function. But is not just that: understanding the connections of these cells may give us a great insight not only in how vision works but also in many conditions including blindness.
So far, more than 80.000 gamers from 130 countries have joined, making EyeWire a great example of "citizen science" and emphasising the significance of not only making science accessible to the community but also of increasing people's participation. Evermore, the 1000s of work-hours spent every day by gamers do not only contribute into the reconstruction of the retinal cells but are also used in order to train computers to process these data more efficiently. The potential of this project is truly full of promises.
You can learn more about EyeWire and play the game here:
Athanasios Alexandris
Azevedo, F. A. et al (2009). Equal numbers of neuronal and nonneuronal cells make the human brain an isometrically scaled-up primate brain. J. Comp. Neurol. 513, 532–541.
Pakkenberg, B., et al (2003). Aging and the human neocortex. Exp. Gerontology, 38:95-99,
Pakkenberg, B.et al. (1997) Neocortical neuron number in humans: effect of sex and age. J. Comp. Neurology, 384:312-320, 1997.