Deep neural networks for the development of anti-aging technologies

Deep neural networks

Neural networks are computer models based on the organization and functioning of neural networks of living organisms. Some people call them Artificial Intelligence, although in reality these are programs capable of self-training strictly within the framework of the questions for which they were made.
Modern processors allow you to create and train fairly complex multi-level neural networks that can, for example, recognize images better than humans and beat people in complex board games..

As part of our project, we will use neural networks to form a biological age estimation algorithm based on the values obtained during a spectrophotometric study.

In the future, the use of a trained neural network will allow:

  • identify the narrow spectral ranges that mostly contribute to the assessment of biological age
  • based on data on narrow ranges, identify biomarkers that mostly contribute to the assessment of biological age
  • create a neural network on human data and use it for individual assessment of biological age
  • create a neural network on the data of laboratory animals (mice, rats, etc.)
  • assess the effectiveness / inefficiency of drugs that are claimed as geroprotectors, since long-term observations will not be necessary to evaluate effectiveness
  • develop a methodology for the selection of individual regimens of geroprotective drugs

Neural networks will be useful in the development of an optimal mode of procedures aimed at combating the aging of the body.