Luminance information decoding on the basis of local field potential signals of pigeon optic tectum neurons

Important aspects of brain information processing can be understood by examining decoding of visual stimuli from neuronal response signals. In this research, the luminance information is decoded from the local field potential signal in the optic tectum region of the pigeon. We designed a luminance visual stimulus model with transient flicker characteristics, recorded multichannel local field potential (LFP) signals using a microelectrode array, extracted LFP Fourier transform energy and phase features, constructed a multivariate linear inverse filter luminance information decoding algorithm, and evaluated decoding effects using a cross-correlation method. We found that LFP signal phase decoding of luminance information yielded better effects than amplitude decoding of luminance information. In the case of optimal frequency band, channels, delay time, and other parameters, the results of phase and amplitude codecoding could reach 0.94±0.02. Comparing the differences between neuronal spike decoding and LFP decoding, we found that LFP signal phase and amplitude codecoding resulted in luminance closer to that of the actual stimulus and required fewer decoding electrode channels.