A viewable three-dimensional map from scientists at Harvard University and Google of one-millionth of the cerebral cortex has appeared on the network. The map was created using 225 million images and 1.4 petabytes of data.
The human brain is the most sophisticated computer ever to exist, and mapping this dense tangle of neurons, synapses, and other cells is nearly impossible. But engineers at Google and Harvard University have created an easy-to-browse and search three-dimensional map of a small area of the human cerebral cortex.
The researchers started with a sample taken from the temporal lobe of the human cerebral cortex, measuring only 1 mm³. It was subdivided into 5,300 slices, each about 30 nanometers (nm) thick. Then the samples were scanned using an electron microscope with a resolution of up to 4 nm. As a result, scientists created 225 million two-dimensional images, which were then combined into a three-dimensional image.
Machine learning algorithms scanned the sample to identify the various cells and structures inside. After several passes through various automated systems, the scientists checked the data to make sure the algorithms were correctly identifying it.
The end result, which Google calls the H01 dataset, is one of the most complete maps of the human brain ever produced. It contains 50,000 cells and 130 million synapses, as well as smaller segments of cells – axons, dendrites. The map is composed of 1.4 petabytes of data – that’s over a million gigabytes.
A companion job preprint is available on the bioRxiv website.