Abstract We have used information theory to analyse the responses of neurons in area 21a of the cat to disparity stimuli. Visual stimuli consisted of drifting sinusoidal grat- ings presented simultaneously to each eye. The relative spatial phase of the gratings varied between stimulus peri- ods in a pseudo-random sequence of 45° increments that covered the full 360°. The mean information content of the responses of all neurons across all phases was 0.72 bits (±0.10, SE, n=29). The information conveyed by each neu- ron was well correlated with the extent to which the inter- ocular phase difference modulated the response of the cell. However, information content was not simply related to firing rate, as there was usually significant information content in the neuronal responses to phase differences that elicited the minimum firing rate. In general, burst respons- es (impulse intervals <4 ms) did not convey more informa- tion than that conveyed by the total response. The contri- bution to the cumulative information of the response in successive 100-ms segments decreased over the course of the 1-s stimulus. The ratio of information transmitted at 200 ms to that transmitted over the full second had a medi- an of 0.30 while the ratio of 500 ms to 1 s was 0.68. Keywords Stereoscopic depth perception · Area 21a · Information theory · Disparity · Interocular phase difference Introduction The visual cortex of cats and primates is suggested to have two largely parallel processing streams for motion and for form (Merigan and Maunsell 1993; DeYoe et al. 1994; Dreher et al. 1996). In the cat, the first region of extrastri- ate cortex on the form pathway is area 21a, which receives input from both areas 17 and 18 but little direct input from the dLGN (Symonds and Rosenquist 1984; Morley et al. 1997). Neurons in area 21a have relatively sharp spatial frequency and orientation selectivity, but show little direc- tional selectivity and have predominantly low pass proper- ties for temporal frequency tuning (Wimborne and Henry 1992; Tardif et al. 1996; Morley and Vickery 1997). The receptive fields of area 21a neurons are mostly binocular, and two-thirds of neurons display profound modulation of their discharge in response to binocular disparity (Wang and Dreher 1996; Vickery and Morley 1999). We have now used information theory to analyse the responses of these neurons to disparity stimuli. This has enabled us to compare the information conveyed by the neuron over various time scales, and by the bursts in the response com- pared with the total response. Materials and methods Extracellular recordings were made from single neurons in area 21a in the anaesthetised adult cat. The procedures for animal care, anaesthesia, surgery and recording, detailed in Vickery and Morley (1999), complied with the guidelines of the National Health and Medical Research Council on the Care and Use of Animals for Scientific Purposes and with those of the Animal Care and Ethics Committee of the University of New South Wales. Visual stimuli consisted of drifting sinusoidal gratings at a Michelson contrast of 0.8 generated using a Cambridge Research Systems video board (Rochester, UK) and presented on two computer monitors (NEC; model XV15) viewed through half-silvered mirrors, which result- ed in a mean luminance of 10 cd/m 2 . In binocular stimulation grat- ings were presented simultaneously to each eye for 1.1 s and grat- ing parameters (spatial and temporal frequency and orientation) were optimal for the dominant eye. The relative phase offset be- tween the gratings was varied between stimulus periods in a pseu- do-random sequence of 45° increments that covered the full 360°. The data was analysed using information theory (Cover and Thomas 1991; Rieke et al. 1997; Buracas et al. 1998), with eight stimulus categories, and response categories equal to the maxi- mum response rate in impulses per second. Each combination of stimulus parameters was presented 5 times; we therefore used gaussian functions based on the mean and SD of the response to the stimulus category rather than the raw data to fill the response categories. In runs with very low response rates, the lower tail of the gaussian called for responses with negative spike counts, and in cases where such negative counts comprised more than 5% of the distribution we fitted a Poisson function to the mean rate. An R.M. Vickery ( ✉ ) · J.W. Morley School of Physiology and Pharmacology, University of New South Wales, Sydney 2052, Australia e-mail: richard.vickery@unsw.edu.au Tel.: +61-2-93851676, Fax: +61-2-93851059 Exp Brain Res (2002) 145:130–132 DOI 10.1007/s00221-002-1108-7 RESEARCH NOTE R. M. Vickery · J. W. Morley Coding of disparity information in extrastriate cortex of the cat Received: 4 October 2000 / Accepted: 8 April 2002 / Published online: 4 May 2002 © Springer-Verlag 2002