What do V1 neurons l ike about natural signals

Sensory systems have been found to adapt their machinery to process natural signals efficiently. What are the statistical features in natural signals that could be driving neuronal adaptation? Our study demonstrates that neural coding of macaque primary visual cortex (area V1) neurons is adaptive to the scale-invariant long-term correlation in the signals. Power spectrum of such signals displays a 1/f power law behavior, which is universal in natural signals. We found that V1 neurons exhibit higher efficiency, transmit more information and carry less noise when responding to 1/f signals than to other kinds of signals. These findings suggest that the appropriate long-term correlation might be a key driving force in shaping and optimizing the machinery of neurons in their adaptation to the natural environment.