New functional neuron classification:

New functional neuron classification:

The discovery of neuron-detectors: the value of the work of Mac-Kolokh, Pits, Maturan and others. They changed the method of studying neurons. We turned to experiments with frogs, where object simulators were used as a stimulus, and not points.

It turned out that if the frog was recorded not even the cortex, but the retina, then there are various neurons: some were excited for the appearance of small spots, others for large spots with a corner. The concept of a neuron detector was introduced - a neuron that registers stimuli with a specific set of properties. Such neurons permeate the entire nervous system, of which the sensory system is built.

Were found stable characteristics of neurons:

- moving edge detectors;

- contrast detectors;

- moving angle detector.

Some neurons responded to the clarity of the border, others to the difference of contrasts, and others to a moving angle.

Research on the cat (its visual cortex) led to the discovery of new types of detectors ( Hubel and Wiesen ). Found neurons that are sensitive to the orientation of visual stimuli (strips that were differently oriented and moved). Neurons were detected that respond to a specific orientation of the strip and move it in one direction. If the strip was moving in a different direction or had a different location, then the neuron was not excited. In addition to such simple neuron detectors, there are more complex ones that are focused on a specific object (for example, on a specific person). A set of answers to the length of a line can be given graphically - few - many - few - no. If it is with a peak, then it is a detector, if without a peak, then it is not a detector, but rather an adder. Neurons are located in the cortex in columns - structural packaging of neurons in the cortex. It is now believed that neurons form a conglomerate in which the bars are arranged in a circle.

Konorsky hypothesized the existence of gnostic neurons - identifying integral objects. Experimental confirmation of these neurons or gestalt units has been obtained. These neurons were found in the associative cortex. There is a gestalt pyramid (Sokolov), which points to the connection of simple neuron detectors and gnostic neurons: simple detectors à complex detectors à gnostic neuron.

A neuron was found near the ram and was excited by the appearance of a shepherd or dog. Consequently, he was more responsive to the function common to the dog and the shepherd, rather than to their appearance.

Rawls worked with monkeys, studies have shown that in the temporal cortex of the monkey, you can find 20% of Gnostic neurons localized in the upper part of the temporal cortex, aimed at identifying:

1 group: a certain familiar person or his photos of people or monkeys.

Group 2: neurons that react not to the face itself, but to the emotion that it expresses.

The properties of gnostic neurons depend on knowledge, that is, they are the product of experience and learning. Detectors are genetically determined. The principle of learning: 1 neuron - 1 person. In addition to the Gnostic neurons, there are subgroups of neurons for which you need a face in face or profile, for others it does not matter, it means there are neurons with a more generalizing function, but there are those with a bounded one. For the complete identification of a person, the work of the entire subgroup of identifying neurons is necessary, and not just one. The more they work, the higher the recognition result. There are neurons that respond to certain gestures.

For identification, the associative cortex must be in activity for a long time. For this, the information is being rewritten in the prefrontal cortex — the same neurons were found there.

Found areas that are activated on non-living objects (buildings, tools, etc.).

Experiment with monkeys. They took 99 unfamiliar images and trained monkeys to identify them. The check was made by pressing the monkey button with a familiar image and reinforcement. After all the images were memorized, electrodes were implanted, which determined the excitation of neurons on a familiar image. When displaying 99 unfamiliar images, not a single neuron was excited.

Now, instead of implanting a neuron trying to use the EEG.

The potential of a local field is a summary characteristic of the potential of a local field. There is a hard connection of the spikes of one neuron with its potential. It is with the local potentials that we encounter at EEG.