Experts from the Southern Federal University Center for Neurotechnology Research and Development (SFedU) have created a mouse-AI biohybrid system that can be used to detect substances in the air. The accuracy of such a system has been increased from 60% to 100%. This was reported to socialbites.ca in the Priority-2030 program of the Ministry of Science and Higher Education of the Russian Federation.
Until now, the olfactory ability for devices has been very limited, as scientists have not yet determined all the components of odorants and the principle of their action on olfactory bulbs. That is why it was decided to combine the electronic nose with the biological one.
In this biohybrid system, the mouse acts as a sensor due to its keen sense of smell, and artificial intelligence analyzes the animal’s brain activity and provides results.
“Microelectrodes are implanted in the olfactory bulb of a mouse. Next, the animal is placed in a state of anesthesia and placed in a box in which a biological signal amplifier is installed. The computer program records the activity of the part of the brain responsible for smell, and artificial neural networks determine the substance presented to the mouse, ” Peter Kosenko, Ph.D.
He noted that depending on the duration of anesthesia, it affects the central nervous system in different ways. The scientists studied the effect of two-hour anesthesia dynamics on the intensity of the rat’s olfactory system and provided a mathematical description of the process.
“During anesthesia, the olfactory brain works much more efficiently than in other states (awake or asleep). We also found that as anesthesia develops, the response of neurons increases and reaches its maximum activity in the second hour, the deepest stage,” said Petr Kosenko.
Such systems can be applied in various fields. For example, they can be used to detect potentially dangerous substances such as explosives or drugs in the air a person breathes, as well as substances specific to oncological diseases.
In the future, experts plan to create systems that will work continuously in real time.