Evidence for neural adaptation to phase spectra induced by the ocular optics

Abstract

Purpose : Our previous work has shown that long-term neural adaptation to poor optics play a critical role in neural processing of both contrast and phase. We hypothesized a mechanism that the visual system can compensate for the optical phase-transfer function imposed by the habitual optics. The goal of this study is to examine evidence for this compensatory mechanism.

Methods : A blur adaptation paradigm was developed to assess shifts in phase perception during short-term exposure (~70mins) to blur induced by a fixed amount of vertical coma (Z^-1₃ = -0.4µm) induced using adaptive optics (AO). Each session consists of three segments: baseline (aberration-free condition under AO correction), blur adaptation (AO-induced vertical coma), and post-adaptation (identical to baseline). Four participants performed the phase perception task during the three parts in which participants judged the appearance (upwards or downwards) of suprathreshold horizontal compound grating stimuli consisting of two sinusoids (frequencies f and 2f) varying in relative phase. The point-of-subjective equality (PSE) was estimated by Weibull psychometric function at different time points. A series of high contrast natural grayscale images and checkerboards was presented between each compound grating and served as perceptual cues for the visual system to detect the presence of blur and induce neural adaptation to blur.

Results : PSE estimates were close to the expected 0° shift during baseline and inducing vertical coma resulted in a significant phase shift relative to baseline (+30.7°), which matched the expected shift in relative phase (+29.5°) from the phase transfer function for -0.4µm vertical coma. Interestingly, the magnitude of the phase shift gradually decreased during blur adaptation, having reduced by 45% (+16.9°) at the end of the blur adaptation period. Moreover, when AO-correction was reengaged (i.e., post-adaptation), PSE did not return to the baseline level. Instead, a significant after-effect in the opposite direction was observed (-14.5°) and matched the attenuation in phase shift during the adaptation period (-13.8°). Following this brief after-effect, PSE returned to the expected baseline level after 2-3 min.

Conclusions : These results show compelling evidence that the human visual system can detect and compensate for disruption of phase congruency via changes in sensory processing over short-term periods of blur adaptation.

This is a 2020 ARVO Annual Meeting abstract.