High-resolution two-dimensional spatial mapping of cat striate cortex using a 100-microelectrode array

Much of our understanding of the visuotopic organization of striate cortex results from single-electrode penetrations and serial recording of receptive field properties. However, the quality of these maps is limited by imprecision in quantifying electrode position, combining data from multiple laminae, and eye drift during the measurement of the receptive field properties. We have addressed these concerns by using an array of 100 closely spaced microelectrodes to investigate the two-dimensional visuotopic organization of layer IV in cat striate cortex. This array allowed simultaneous measurement of the receptive field properties of multiple single units on a regularly spaced grid. We found the relationship between cortical and visual space to be locally non-conformal: the receptive field locations associated with a closely spaced line of electrodes appeared randomly scattered in visual space. To quantify the scatter, we fitted a linear transformation of electrode sites onto the associated receptive field locations. We found that the distribution of the difference between the predicted receptive field location and the measured location had standard deviations of 0.59 degrees and 0.45 degrees in the horizontal and the vertical axes, respectively. Although individual receptive field positions differed from the predicted locations in a non-conformal sense, the trend across multiple receptive fields followed the maps described elsewhere. We found, on average, that the 13 mm2 of cortex sampled by the array mapped onto a 5.8-degrees) region of visual space. From the scaling of this map and a combination of the statistics of the receptive field size (2.7+/-1.5 degrees) and scatter, we have explored the impact of electrode spacing on the completeness and redundancy in coverage of visual space sampled by an array. The simulation indicated an array with 1.2-mm spacing would completely sample the region of visual space addressed by the array. These results have implications for neuroprosthetic applications. Assuming phosphene organization resembles the visuotopic organization, remapping of visual space may be necessary to accommodate the scatter in phosphene locations.