Scientists at the St. Petersburg Federal Research Center of the Russian Academy of Sciences have created an open dataset that will help assess a person's level of concentration on a task. It is based on indicators of brain activity, heart rate, head and chest movements.
Developers will be able to use this data to train neural networks to distinguish between states of concentration and distraction. In the future, such models could be used in applications for attention training, including for children with attention deficit hyperactivity disorder (ADHD).
The obtained data can be used to train a neural network: give the algorithm the task of classifying a person's concentration level based on EEG and ECG. Further, based on such a model, programmers will be able to create various applications, for example, for attention training in children with attention deficit disorder.
33 volunteers participated in the study. Fifteen of them regularly practiced yoga and meditation for at least a year, while the rest had never tried such practices. During the 45-minute experiment, participants performed five tasks – from a relaxed state and free “mind wandering” to concentrating on an imaginary point and searching for objects in pictures.
Analysis showed that in people with experience in yoga and meditation, transitions between focus and distraction are more noticeably reflected in EEG and ECG indicators. One of the key markers of concentration was the signals from the prefrontal cortex, which is responsible for attention.
In the future, researchers want to try to train neural networks to determine the level of concentration not only from EEG and ECG data, but also from video images of a person. The project was carried out with the participation of HSE University in St. Petersburg, RUDN University, and MISIS University with the support of the Russian Science Foundation.
Read more on the topic:
- Software for controlling electronics with gestures with high accuracy developed at St. Petersburg Federal Research Center of the Russian Academy of Sciences
- MASAI Emotion Recognition System Based on Voice and Facial Expressions Developed by Scientists at the St. Petersburg Federal Research Center of the Russian Academy of Sciences