TY - JOUR T1 - Cortical Neuroprosthesis Merges Visible And Invisible Light Without Impairing Native Sensory Function JF - eneuro JO - eNeuro DO - 10.1523/ENEURO.0262-17.2017 SP - ENEURO.0262-17.2017 AU - Eric Thomson AU - Ivan Zea AU - William Windham AU - Yohann Thenaisie AU - Cameron Walker AU - Jason Pedowitz AU - Wendy França AU - Ana L. Graneiro AU - Miguel A. Nicolelis Y1 - 2017/12/07 UR - http://www.eneuro.org/content/early/2017/12/07/ENEURO.0262-17.2017.abstract N2 - Adult rats equipped with a sensory prosthesis, which transduced infrared (IR) signals into electrical signals delivered to somatosensory cortex (S1), took approximately 4 days to learn a four-choice IR discrimination task. Here, we show that when such IR signals are projected to the primary visual cortex (V1), rats that are pre-trained in a visual-discrimination task typically learn the same IR discrimination task on their first day of training. However, without prior training on a visual discrimination task, the learning rates for S1- and V1-implanted animals converged, suggesting there is no intrinsic difference in learning rate between the two areas. We also discovered that animals were able to integrate IR information into the ongoing visual processing stream in V1, performing a visual-IR integration task in which they had to combine IR and visual information. Furthermore, when the IR prosthesis was implanted in S1, rats showed no impairment in their ability to use their whiskers to perform a tactile discrimination task. Instead, in some rats, this ability was actually enhanced. Cumulatively, these findings suggest that cortical sensory neuroprostheses can rapidly augment the representational scope of primary sensory areas, integrating novel sources of information into ongoing processing while incurring minimal loss of native function.Significance Statement Using a sensory neuroprosthesis that projects information from the IR environment to primary sensory areas, we show that adult rats can rapidly integrate completely novel sensory information into pre-existing cortical maps. When the prosthesis is implanted in V1, animals learn to perform a multimodal integration task, fusing IR and visual information that is simultaneously superimposed on the same cortical area. When the prosthesis is implanted in S1, the tactile function of S1 is left undisturbed, and often enhanced. Hence, it is possible to merge multiple streams of information onto the same primary cortical area without compromising its original function. This is auspicious for the development of sensory prosthetic systems for adult victims of brain injury. ER -