Analysis of Mouse Motor Activity in Free Behavior by Bioradiolocation

Currently, there are no inexpensive methods for long-term non-contact assessment of the motor activity of small laboratory animals in free behavior, so the development of new reliable methods is an urgent task of modern medicine. The paper proposes a method for registering motor activity based on the bioradiolocation principles, which allows remotely registering the vital signs of an animal determined by the movements of the surface of the animal's body during movement, breathing and heartbeat. In particular, this paper proposes an algorithm for classifying episodes of mouse motor activity, the effectiveness of which was evaluated on the base of the Institute of Higher Nervous Activity and Neurophysiology of RAS (IHNA and NPh RAS, Russia, Moscow). The algorithm is based on the k-nearest neighbors method, which uses as a feature vector a set of 8 characteristics of the bioradiolocation signal, the choice of which is justified in the work. This method classifies experimental data with a sensitivity of 97 % and a specificity of 99 %.