A Trip Through a 3D-Modeled Brain

See how a mouse's mind sees something.

Brains are, by design, incredibly dense. Whether a particular brain belongs to a human or a mouse, it features layer upon layer of matter that twists and turns and is almost incomprehensible in its complexity.

Our minds are a little bit mind-boggling.

Which is what makes the video above, produced by Nature, so remarkable. It depicts a 3D model of a little piece of a mouse's brain -- specifically, its retina. And the model is, save for its rainbow-bright colors, a faithful one: It shows the precise shape and location of 950 different cells, including both cell bodies and cell branches (dendrites). Each color represents the branches of a different nerve cell.

So how do you make a model like this? First, you take a mouse's retina and slice it extremely thinly. Then you put those slices under an electron microscope. Then you analyze the images using a computer, assigning different colors to each structure. From there, you let your computer take over. In this case, researchers' software was able to identify not just the shape, but also the volume of each structure in the mouse's retina -- and to identify the synapses, or points at which the different cells connect to each other.

That was only the beginning, though. Some 300 students -- human ones -- also worked on the mouse-mind-mapping project, spending a total of 30,000 hours (!) tracing the path of each neuron. Which allowed the computer to know, in turn, how to fuse the colored sections in its visualization.

Here's something else that's remarkable: All that effort -- and the intricate model that resulted from it -- represents only 0.06 percent of the mouse's retina. Imagine what it would take, the video notes, to map an entire mouse brain ... not to mention a human one, which contains, in general, more than 80 billion neurons.

So, largely because of the resources it requires, the mapping approach has been controversial. Some argue, Nature notes, that the technique's returns -- a visual model -- don't justify their expense. Nature points out, however, that one brain model recently led to the discovery of a new form of cell in the mouse retina, the XBC -- a bipolar cell that relays visual information from photo receptors to other cells in the network, and a cell that had formerly evaded human detection. Scientists have also used mapping to understand the brains of fruit flies. The studies, Nature argues, "demonstrate that extraordinarily detailed anatomy, combined with physiology, is furthering our understanding of the brain."

So could we apply that understanding, and these modeling techniques, to the human brain? Not yet. (Again, 30,000 hours of work to map 0.06 percent of a brain. Of a mouse.) But, then again, we know the technology we need to get there. "Enthusiasts argue," Nature puts it, "that if you really want to understand the brain, you'd better get mapping."