Scientists using a 16-beam two-photon microscope found that the noise bursts, which are born simultaneously in different neurons, restrict coding of information in populations of thousands or more cells in the visual cortex of mice. The strongest kind of noise do not overlap with the activity that was associated with touch stimuli. Thus, the system protects a signal from the main part of interference, a limiting factor is the weaker noise activity. Article published in the journal Nature.
Nerve cells have spontaneous activity, and often for coding signals they use electrical impulses themselves, and the change in the frequency of their occurrence. In this case, the noise interferes with neurons to detect weak stimuli; perhaps that is it limits the precision of information coding in nerve cells. However, the perception of incentives is based on the work of entire neural networks, not individual cells. The averaging of signals from different cells may reduce the amplitude of the noise and help bring out a clean signal. Or Vice versa — if the noise of individual neurons with similar characteristics occur simultaneously, the desired signal is more damped.
Previously investigated only the noise that occurs in pairs of cells in such experiments is high measurement error. Oleg Rumyantsev from Stanford University and his colleagues recorded the activity of entire neural populations, to highlight patterns of noise and to estimate its influence on the accuracy of coding. To do this, scientists have created a two-photon microscope, in which 16 laser beams scanned region of the visual cortex of mice with an area of four square millimeters.
The researchers observed the activity of five neurons of mice that were presented with visual stimuli (oblique contrast stripes). This used calcium imaging, in which the label is associated with the free calcium in the cytoplasm and begins to fluoresce. The brighter the fluorescence, the higher the calcium concentration — a clear indicator of the activity of nerve cells. Each animal was able to simultaneously observe one or two thousand neurons.