Purchase this article with an account.
Nikolay Akimov, Rene C Renteria; Comparison of retinal ganglion cell receptive field properties measured using Gaussian and binary (non-Gaussian) white noise checkerboard stimuli. Invest. Ophthalmol. Vis. Sci. 2015;56(7 ):3244.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
White-noise checkerboard mapping is a valuable method for characterizing the receptive fields (RF) of retinal ganglion cells (RGCs). Pseudo-random, binary (black-and-white) checkerboard stimuli can achieve higher contrast than Gaussian white noise stimuli and so can provide receptive field mapping in less time or be presented at higher frequencies to achieve higher temporal resolution. Here, we compared receptive field sizes determined from stimulation using Gaussian and binary white noise checkerboards.
Retinas from 4 C57BL/6J adult mice were isolated and mounted on a multi-electrode array (60 electrodes, 100 µm separation; MultiChannel Systems). Retinas were perfused with Ames’ medium at 320C and stimulated using an OLED screen (eMagin, Inc). After adaptation to the mean stimulus intensity, action potentials were recorded during presentation of two sets of checkerboard images (48 μm check size), each lasting 30 min. The intensity distribution of checks was Gaussian and presented at 30 Hz for the first set and binary (black and white) and presented at 60 Hz for the second set. The most effective stimulus was determined by averaging the images which were presented in the 1 sec preceding each spike (the spike-triggered average or STA). A 2D Gaussian fit to the peak frame of the STA was used as a spatial RF map representation. Map quality was measured as the ratio of maximum STA signal and STA standard deviation (SD). RGC RF diameters were measured at 1 SD of the 2D Gaussian fit.
ON RGC RF diameters, when measured using either Gaussian or binary white noise stimuli, were the same (Gaussian 120±3 μm, binary 118±4 μm; mean±SEM, N=42). OFF RGC RF diameters were significantly larger (by 17%) when measured with binary compared to Gaussian white noise (Gaussian 123±5 μm, binary 143±6 μm, N=29). Map quality time-course analysis showed that 95% of the RGCs could be mapped in 10.5 min using Gaussian and 6.6 min using binary white noise stimuli.
For analysis of RGC receptive field properties, binary white noise stimuli can be presented at higher refresh rates than Gaussian white noise stimuli to achieve higher temporal resolution with shorter presentation times. Care must be used when comparing receptive field sizes measured with these two methods.
This PDF is available to Subscribers Only