|
Listen : Audio version of this article
|
The human brain consists of 86 million nerve cells and more than a trillion connections between each one. No doctor, no matter how high his qualifications are, cannot recognize a nerve cell without at least years of study under the microscope, but there are still chances that any scientist may fail with the last set of data.
Scientists from the Max Plank Institute of neurobiology and Google AI have found the solution to this complex problem, by creating and training artificial intelligence on neurobiological analyses that can recognize and divide the nerve cells, separately, just based on their appearances. While scientists have used serial block-face scanning electron microscopy, the brain cells and connections are all automatically surveyed and displayed in the form of a three-dimensional image, that helps the AI recognize the cells easily.
The data read by the ai can easily be transferred into any computer, that can be further processed into data sets. “It will take many months to survey a zero.3 mm3 piece of the brain below AN microscope,” says Philipp composer, degree student in Winfried Denk’s Department at the Max Planck Institute of neurobiology.
“Depending on the scale of the brain, this appears like loads of your time for a little piece. however, even this contains thousands of cells.” Such a knowledge set would additionally need nearly a hundred terabytes of cupboard space. However, it’s not the gathering and storage but rather the data analysis that’s the troublesome half.
Similar algorithms and AI can also be used to know where a certain nerve belongs to in a human body and can give a visual process on the nerves. This can assist in the whole medical profession, from surgeries to basic medication that requires data.