Scientists proposed a method for predicting the level of intelligence of children on MRI of the brain using three-dimensional depth training networks.
Magnetic tomography is a common way to obtain pictures of human internal organs. The researchers decided to explore the ability to use the images of the brain thus obtained to determine the biological intellectual potential of adolescents, regardless of the level of knowledge, age and gender.
To do this, the team from ScoThech created the architecture of a three-dimensional neural network, which to improve the accuracy of the forecast applied to the same data at once several mathematical models. To analyze MRI with the help of engine vision, a new ensemble method was used, which allows you to classify the entire snapshots and without losing information.
In the course of the study, the correlation was discovered between the level of the «teaching» intelligence of the child and the anatomy of his brain. Although the prediction accuracy is still below the desired level, but the developed models open up new aspects of children’s development.
Machine training technologies also allow to identify signs of children’s disease, which even doctors may not pay attention. For example, recently researchers
To develop the channel, your support is important to us, subscribe to the channel and put like.