Preparation of the data for further CNN training to atomized and optimized histological researches

Apr 16, 2019, 5:00 PM
Cafe (House of Scientists)


House of Scientists

Poster Information Technology Poster session


Inna Kolesnikova (by friend)


This work aims to adapt convolutional neural network for neuromorphological researches of brain tissue. The standard stain Nissle method allows identifying different types of brain cells with damages. The stained histologic sections are investigated for damages in ones using optical microscope. It is important to bear in mind that this research method is rather judgmental and energy intense. Therefore, this work consists analysis free and existing services/platforms that allow studding neural network. Because of such factors as personality trait, physiological make-up of every pathologist it may happen that the neuron would be identified wrong. Therefore, the human factor is important and may have a prominent influence. Consequently, the project is currently important. By this research it is made the preparation data for further recognition of convolutional neural network. The development of such programme code of detection brain cells allows cutting research time shorter, the method is more objective, the calculating is more correct. A faster and more accurate analysis speeds up the process of obtaining the final result in the study of the effect of any factor on the brain structure.

Primary author

Inna Kolesnikova (by friend)

Presentation materials

There are no materials yet.