Reverse engineering the brain, one slice at a time

Narayanan Kasthuri and a team of researchers  from Harvard, MIT, Duke, and John Hopkins universities, fully reconstructed all the neuron sections and many sub-cellular objects, including synapses and synapse vesicles, in a volume of 1500 µm3 (which is just a little more than one millionth of a cubic millimeter) using 3×3×30 nm voxels.

cilinder

The results, published in an article in the journal Cell, in July 2015, describe the experimental procedure and the conclusions. The data was obtained by collecting 2,250 brain slices, each roughly 30 nm thick, obtained with a tape-collecting ultramicrotome that slices brain sections using a diamond knife.The slices were imaged using serial electron microscopy and the images processed in order to reconstruct a number of volumes. In this volume, the authors have reconstructed the 3D structure of the 1500 µm3 of neural tissue, which included hundreds of dendrites, more than 1400 neuron axons and 1700 synapses, which corresponds to about one synapse per cubic micron.

(Rendering by the authors, used with permission)

 

Advertisements

Computers finally excel at Go

 

game_of_go

Go is a beautiful game, with a very large branching factor that makes it extremely hard for computers. For decades, playing this game well was outside the reach of existing programs.

We just learned that computers finally mastered Go, in a paper published in the journal Nature. By using machine learning techniques and, in particular, deep learning, the program AlphaGo, created by Google’s company DeepMind, managed to beat Fan Hui, the European Go champion, five times out of five. Whether AlphaGo is sufficiently strong to beat the best players in the world, remains to be seen. However, it already represents a very significant advance of the state of the art.

What was maybe the last bastion in table games still unconquered by computers is no more. Computers are now better than humans at all table games invented by humanity.