The neural network was taught to search and analyze nanoparticles through a microscope

In order to teach a neural network to perform this task, it needs to be shown several tens of thousands

tagged photos. These are special images that show what task the neural network will perform. This, according to scientists, is extremely difficult for a number of highly specialized scientific tasks.

To get around this problem, scientists did not tag real photographs to train the neural network, but generated images that simulate them on a computer.

SEM (scanning electron microscope)Using a beam of electrons instead of visible light, it is used in the study of nanoparticles synthesized for medicine and other purposes. Analysis of SEM images consists of detecting particles and distributing them by size. Neural network approaches in this area are not developed, and standard image processing methods do not provide the required quality. 

Alexander Kharin, specialist at the Engineering Physics Institute of Biomedicine, National Research Nuclear University MEPhI

The results of the study will make it possible to automate the processing of SEM images, revolutionizing standard methods for studying new materials, scientists are confident.

This work, the authors believe, will help not onlyto reduce the research time, but also to increase the number of analyzed particles - from hundreds of units to tens of thousands. In the future, the research team is going to automate the classification of nanoparticles.

Read also

A meteorite fell on an Indonesian house. Its owner became a millionaire

Neurons in the human brain and the network of galaxies are similar

Due to plate movement, the Pacific Ocean floor is now deep under China