Efficient Coding of Natural Scenes in the Lateral Geniculate Nucleus: Experimental Test of a Computational Theory

Abstract

A recent computational theory suggests that visual processing in the retina and the lateral geniculate nucleus (LGN) serves to recode information into an efficient form (Atick and Redlich, 1990). Information theoretic analysis showed that the representation of visual information at the level of the photoreceptors is inefficient, primarily attributable to a high degree of spatial and temporal correlation in natural scenes. It was predicted, therefore, that the retina and the LGN should recode this signal into a decorrelated form or, equivalently, into a signal with a "white" spatial and temporal power spectrum. In the present study, we tested directly the prediction that visual processing at the level of the LGN temporarily whitens the natural visual input. We recorded the responses of individual neurons in the LGN of the cat to natural, time-varying images (movies) and, as a control, to white-noise stimuli. Although there is substantial temporal correlation in natural inputs (Dong and Atick, 1995b), we found that the power spectra of LGN responses were essentially white. Between 3 and 15 Hz, the power of the responses had an average variation of only +/-10.3%. Thus, the signals that the LGN relays to visual cortex are temporarily decorrelated. Furthermore, the responses of X-cells to natural inputs can be well predicted from their responses to white-noise inputs. We therefore conclude that whitening of natural inputs can be explained largely by the linear filtering properties (Enroth-Cugell and Robson, 1966). Our results suggest that the early visual pathway is well adapted for efficient coding of information in the natural visual environment, in agreement with the prediction of the computational theory.

Power-law statistics and universal scaling in the absence of criticality

“…In natural environments, regularity of sensory input to the brain may create strong and long-range correlations in space and time [44][45][46][47]. It soon appeared that these correlations would be detrimental for the brain to encode sensory stimuli and detect changes efficiently [48,49].…”

On the number of states of the neuronal sources

“…The time interval [0, T ] is then split in n bins Dt i (1 0/i 0/n ) of the same length. If there are N k spikes in the bin Dt k , then assign the number N k to Dt k (Dan et al, 1996;Rieke et al, 1998;Zador, 1998). The result is a message of length n with no more than n different letters.…”

Neural coding of naturalistic motion stimuli

“…To meet the challenge we must modify both our experimental designs and our methods for analyzing the responses to these much more complicated inputs. Recent examples of laboratory based approaches to the problem of natural stimulation are studies of bullfrog auditory neurons responding to synthesized frog calls (Rieke et al , 1995), insect olfactory neurons responding to odour plumes (Vickers et al , 2001), cat LGN cells responding to movies (Dan et al , 1996, Stanley et al , 1999, primate visual cortical cells during free viewing of natural images (Gallant et al , 1998, Vinje andGallant, 2000), auditory neurons in song birds stimulated by song and song-like signals (Theunissen and Doupe, 1998, Theunissen et al , 2000, the responses in cat auditory cortex to signals with naturalistic statistical properties (Rotman et al , 1999), and motion sensitive cells in the fly Egelhaaf, 2001, de Ruyter van Steveninck et al , 2001). In each case compromises are struck between well controlled stimuli with understandable statistical properties and the fully natural case.…”