Researchers reveal how neurons support flexible decision-making
Beijing (ANTARA) — A team of Chinese researchers has uncovered how neurons in the brain support flexible decision-making.
According to a study recently published in the journal Science, information about rules and sensory information are processed in the same neurons, which also contribute to cognitive computation.
Traditional theory describes neurons as the basic structural and functional units of the nervous system. For a long time, most studies treated each neuron as a simple, point-like unit that performs summation.
However, research in neuroscience and artificial neural networks has found that higher cognitive functions, such as flexible decision-making, require the extensive integration of contextual information and abstract rules, as well as the dynamic adjustment of outputs. Such complexity is difficult to explain using a simple, point-like model of the neuron.
Xu Ninglong, a researcher at the University of the Chinese Academy of Sciences, and his team designed a series of experiments in which mice heard different sounds and had to infer hidden rules. After the mice had learnt the rules, the team changed them to observe how quickly the animals adapted to the new logic. The researchers used a two-photon microscope to observe neurons in the mice’s auditory cortex.
A neuron can be likened to a tree with many branches. The cell body is the main part of the neuron, while dendrites are branch-like extensions. Those farther away are known as apical dendrites.
The experiments found a clear division of labour within the same neuron. Apical dendrites mainly encoded the abstract rules inferred by the mice, while the cell body more strongly represented external auditory sensory input. The two types of signals underwent nonlinear integration within the neuron, producing a decision output that drove behavioural choices.
The experiments showed that a single neuron can function as a microcomputer, and that interactions between its dendrites and cell body can perform cognitive computations associated with decision-making. Only after a hypothesis had been verified repeatedly and its reliability confirmed did the brain gradually strengthen the neuron’s synapses and store the rule permanently.
The team found that although artificial intelligence (AI) can also use context and available information to adjust its responses, its workings differ greatly from those of the brain. Large models generally develop capabilities through training on data, whereas the brain combines learning from experience, internal states and real-time information processing. Even a single neuron has a complex internal computational structure.
The researchers also found that the brain can support flexible behaviour while using limited energy. Studying this biological mechanism could provide insights to help make AI systems more flexible and energy-efficient.