Artificial intelligence scientists from Google and Harvard University (USA) have developed an interactive map of the human brain. On the platform it is possible to visualize thousands of neurons and neuron fragments, 130 million synapses and even cellular and subcellular structures – all with the digital reconstruction of a cubic millimeter of brain tissue.
The resolution of the images is so high that all of this data takes up 1.4 petabytes of space, which is almost 3,000 times the storage capacity of a smartphone with 512 gigabytes of storage.
Images can be freely accessed on a platform called Neuroglancer. On the official website of the project, it is possible to see certain parts of the brain, which makes it easier to use the tool and, for example, in education.
To create the map, the researchers used brain tissue donated from a 45-year-old patient who had undergone surgery to remove an epileptic focus in the temporal lobe on the side of the brain. When removing the affected part, doctors also had to remove healthy tissue from the cortex, which was treated so that it could be cut into more than 5,000 wafer-thin slices (30 nanometers each). These parts were digitized to create the map in three dimensions (3D).
According to Google, this is the largest digitally reconstructed brain tissue sample of all types to date.
According to the company, the main goals of the research are to find new sources for studying the human brain and to promote improvements in connectomics technologies, an area of science dedicated to creating maps and depictions of the connections of the nervous system.
“The size of the data set provides the basis for long-term studies by researchers of the human cortex,” write Tim Blakely and Michael Januszewski, the Google software engineers involved in the project, in a text published by the company.
The researchers also published a pre-print article (not yet reviewed by other scientists) that described the processes that led to the creation of the map. The article also shows that Neuroglancer was used to discover new cell types in the brain, but much of the file has yet to be studied by scientists.