In the brain, as in business, success has a lot to do with making the right connections. In fact, it’s not a stretch to say that our subjective perceptions of ourselves, of one another and of the world around us all come down to how our neurons are connected. Yet the awesome complexity of these connections remains an imposing barrier to scientists who are trying to get at how the mind works by mapping the brain’s circuitry. Now, the Digital Reconstruction of Axonal and Dendritic Morphology (Diadem) Challenge, which kicked off 29 August at the Janelia Farm Research Campus in Virginia, aims to simplify things.
From 125 teams that signed on to the challenge back in April 2009, the field has been winnowed down to five teams which are competing this week in the final round of the competition. Their objective is to create a computer algorithm that can trace a single neuron based on microscope images. The winning team, which will be announced on 1 September, will receive a $75,000 prize.
The organizers of this challenge hope that creating detailed neural maps will allow researchers to gain insight into disorders in the brain and assess the effectiveness of different drugs.
The challenge is daunting. Currently, researchers must sift through images of endlessly-forking neurons and manually trace each branching path. Adding up all the axons (the transmitting end of a neuron), and dendrites, (the receiving ends), might take weeks or months. In a single mouse brain, for example, the combined total length of all the neural branches roughly equals 24 million kilometres, equivalent to all the paved roads in the world combined.
Over the coming days, the five teams in the final found of the Diadem challenge will each try to create an algorithm that can trace a neuron 20 times faster than is now possible. Judges have arranged datasets for the teams to work with and the winning algorithm must be both swift and accurate.
Ready, set, map!
Image courtesy Wikimedia