Purchase this article with an account.
Rebecca Esquenazi, Kimberly Meier, Michael Beyeler, Geoffrey Boynton, Ione Fine; Learning to see again: Perceptual learning for sight restoration technologies. Journal of Vision 2019;19(15):36. doi: https://doi.org/10.1167/19.15.36.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Electronic retinal prostheses, the only sight restoration technology currently approved for implantation in patients, cause simultaneous, rather than complementary firing within on- and off-center retinal cells. Here, using ‘virtual patients’ - sighted individuals viewing distorted input - we examine whether plasticity might compensate for the resulting neuronal population coding distortions. Four participants were trained in a perceptual learning task using dichoptic presentation of original and contrast-reversed images as a proxy for inappropriate on and off-cell stimulation. Each greyscale image (I) and its contrast-reverse (I′) was filtered using a radial checkerboard in Fourier space (F) and its inverse (F'). [I * F?] + [I′* F] was presented to one eye, [I * F] + [I' * F?] to the other, such that regions producing on-responses in one eye produce off-responses in the other eye.
Participants viewed these distorted scenes which had a 50% chance of containing an embedded object in a random size, location, and orientation. Participants indicated whether a cued object was present or absent, and repeated the task once daily for up to 20 one-hour sessions. All participants initially showed poor performance, but continuously improved across each session, as measured by d' scores. Future studies will examine transfer of learning (e.g. 1/f noise in each image pair, monocular presentation, and switching the Fourier filters across eyes) to examine whether participants have learned to suppress or ‘flip the perceptual sign’ of neural populations associated with the contrast reversed image, or have learned how to more efficiently extract visual information from distorted images.
This PDF is available to Subscribers Only