A neural circuit is a population of neurons interconnected by synapses to carry out a specific function when activated.  Neural circuits interconnect to one another to form large scale brain networks. Biological neural networks have inspired the design of artificial neural networks, but artificial neural networks are usually not strict copies of their biological counterparts.

Circuitry

An example of a neural circuit is the trisynaptic circuit in the hippocampus. Another is the Papez circuit linking the hypothalamus to the limbic lobe. There are several neural circuits in the cortico-basal ganglia-thalamo-cortical loop. These circuits carry information between the cortex, basal ganglia, thalamus, and back to the cortex. The largest structure within the basal ganglia, the striatum, is seen as having its own internal microcircuitry.

Neural circuits in the spinal cord called central pattern generators are responsible for controlling motor instructions involved in rhythmic behaviours. Rhythmic behaviours include walking, urination, and ejaculation. The central pattern generators are made up of different groups of spinal interneurons.

There are four principal types of neural circuits that are responsible for a broad scope of neural functions. These circuits are a diverging circuit, a converging circuit, a reverberating circuit, and a parallel after-discharge circuit.

In a diverging circuit, one neuron synapses with a number of postsynaptic cells. Each of these may synapse with many more making it possible for one neuron to stimulate up to thousands of cells. This is exemplified in the way that thousands of muscle fibers can be stimulated from the initial input from a single motor neuron.

In a converging circuit, inputs from many sources are converged into one output, affecting just one neuron or a neuron pool. This type of circuit is exemplified in the respiratory center of the brainstem, which responds to a number of inputs from different sources by giving out an appropriate breathing pattern.

A reverberating circuit produces a repetitive output. In a signalling procedure from one neuron to another in a linear sequence, one of the neurons may send a signal back to initiating neuron. Each time that the first neuron fires, the other neuron further down the sequence fires again sending it back to the source. This restimulates the first neuron and also allows the path of transmission to continue to its output. A resulting repetitive pattern is the outcome that only stops if one or more of the synapses fail, or if an inhibitory feed from another source causes it to stop. This type of reverberating circuit is found in the respiratory center that sends signals to the respiratory muscles, causing inhalation. When the circuit is interrupted by an inhibitory signal the muscles relax causing exhalation. This type of circuit may play a part in epileptic seizures.

In a parallel after-discharge circuit, a neuron inputs to several chains of neurons. Each chain is made up of a different number of neurons but their signals converge onto one output neuron. Each synapse in the circuit acts to delay the signal by about 0.5 msec so that the more synapses there are will produce a longer delay to the output neuron. After the input has stopped, the output will go on firing for some time. This type of circuit does not have a feedback loop as does the reverberating circuit. Continued firing after the stimulus has stopped is called after-discharge. This circuit type is found in the reflex arcs of certain reflexes.

Study methods

Different neuroimaging techniques have been developed to investigate the activity of neural circuits and networks. The use of "brain scanners" or functional neuroimaging to investigate the structure or function of the brain is common, either as simply a way of better assessing brain injury with high resolution pictures, or by examining the relative activations of different brain areas. Such technologies may include functional magnetic resonance imaging (fMRI), brain positron emission tomography (brain PET), and computed axial tomography (CAT) scans. Functional neuroimaging uses specific brain imaging technologies to take scans from the brain, usually when a person is doing a particular task, in an attempt to understand how the activation of particular brain areas is related to the task. In functional neuroimaging, especially fMRI, which measures hemodynamic activity (using BOLD-contrast imaging) which is closely linked to neural activity, PET, and electroencephalography (EEG) is used.

Connectionist models serve as a test platform for different hypotheses of representation, information processing, and signal transmission. Lesioning studies in such models, e.g. artificial neural networks, where parts of the nodes are deliberately destroyed to see how the network performs, can also yield important insights in the working of several cell assemblies. Similarly, simulations of dysfunctional neurotransmitters in neurological conditions (e.g., dopamine in the basal ganglia of Parkinson's patients) can yield insights into the underlying mechanisms for patterns of cognitive deficits observed in the particular patient group. Predictions from these models can be tested in patients or via pharmacological manipulations, and these studies can in turn be used to inform the models, making the process iterative.

The modern balance between the connectionist approach and the single-cell approach in neurobiology has been achieved through a lengthy discussion. In 1972, Barlow announced the single neuron revolution: "our perceptions are caused by the activity of a rather small number of neurons selected from a very large population of predominantly silent cells." This approach was stimulated by the idea of grandmother cell put forward two years earlier. Barlow formulated "five dogmas" of neuron doctrine. Recent studies of 'grandmother cell' and sparse coding phenomena develop and modify these ideas. The single cell experiments used intracranial electrodes in the medial temporal lobe (the hippocampus and surrounding cortex). Modern development of concentration of measure theory (stochastic separation theorems) with applications to artificial neural networks give mathematical background to unexpected effectiveness of small neural ensembles in high-dimensional brain.

Clinical significance

Sometimes neural circuitries can become pathological and cause problems such as in Parkinson's disease when the basal ganglia are involved. Problems in the Papez circuit can also give rise to a number of neurodegenerative disorders including Parkinson's.

Media contacts,
Augustina
Managing Editor
Journal of Brain Research