Commentary: Neural Codes and Information
What is more important - signal or noise?

Alexander Nezhinsky
OmniResearch, Inc.

   Article

The long-held notion that signals are generally more relevant than the background noise has been recently challenged. This notion has its roots in the nearly three centuries of the scientific practice. During this entire period scientists and engineers have been trying obsessively to identify, quantify and qualify various signals separating them from the so called "noise". A new advanced study seems to show that the reverse appears to be definitely right. Three adult specimens of Mexican bean beetle larvae Fig. 1 maintained in perfect physical and mental condition, were subjected to a long series of meticulously performed experiments leading to the revolutionizing results. Prior to the experiments a long record had been taken from a feline primary somatosensory cortex and a standard signal-noise separation procedure was performed (SNR=0.8, P=0.95, T=36.7, a+b=3, a=1, b=?, N> M). Then the following experiments were performed. The "signal", the "noise" and the "raw" waveforms (i.e. signal plus noise, mixed case), after being properly digitized, low-pass filtered and converted back to the analog form, were injected to the alert larvae. After injection of the "signal" waveform all three larvae slowed down their usual activities appreciably and consequently stopped feeding till the end of the experiment. In the mixed case there was no apparent change in activity, while in the "noisy" case all three larvae have increased their rate of the leaf consumption (assessed using the DC component of Fast Fourier Transform performed upon the remaining leaf area time dependence curve). These results indicate unequivocally that the noise waveform was more beneficial to the larvae from the specimen fitness standpoint. Thus, pure noise has much more biological relevance than a pure signal or a common mix thereof. This leads to an astonishing hypothesis that cats brain cells, too, prefer to communicate almost exclusively through noise. It is well known that monkeys are like cats to almost all practical purposes so it can be inferred rather safely that noise is the main medium of communication in the central nervous system of all primates. Taking into account that humans are also primates, this study in a way reinforces the point, too.