I’m no neuroscientist, and yet, here I am at my computer attempting to reconstruct a neural circuit of a mouse’s retina. It’s not quite as difficult and definitely not as boring as it sounds. In fact, it’s actually pretty fun, which is a good thing considering I’m playing a videogame.
Called EyeWire, the browser-based game asks players to map the connections between retinal neurons by coloring in 3-D slices of the brain. Much like any other game out there, being good at EyeWire earns you points, but the difference is that the data you produce during gameplay doesn’t just get you on a leader board—it’s actually used by scientists to build a better picture of the human brain.
Created by neuroscientist Sebastian Seung’s lab at MIT, EyeWire basically gamifies the professional research Seung and his collaborators do on a daily basis. Seung is studying the connectome, the hyper-complex tangle of connections among neurons in the brain.
More specifically, right now Seung and his researchers are reconstructing neural circuits in the retina to get a better idea of how humans perceive directional motions. So for example, when you see something that causes you to look up or down, it’s believed that there are certain cells that respond to that stimuli.
Seung and his group aren’t exactly sure how those cells work or how they’re connected to one another, but they’re trying to find out. “Sebastian likes to say, ‘If we don’t understand how something as simple as motion perception works, how are we going to be able to answer these higher level questions like what happens in learning or mental disorders?’” says Amy Robinson, creative director of EyeWire. The brain is science’s version of the wild west—it’s basically uncharted territory that is waiting to be explored and discovered.
But even figuring out something as seemingly simple as motion perception is incredibly time consuming and has led to a huge bottleneck in the lab’s research around image analysis. Robinson says it currently takes the lab around 50 hours to reconstruct one neuron, even with help from artificial intelligence. Multiply that by the 85 billion (the approximate number of neurons in a human brain), and you can see how they might need some help.