Visual Prostheses—Principle and Technology from Biomimetics Perspective

Research on visual prostheses, which transmits visual information to the brain by artificially stimulating the remaining visual nervous system in patients with acquired blindness, is underway. Generally, visual prostheses aim to restore the function of damaged neural tissue due to disease, by sending visual information to the higher area of the visual system usually through electrical stimulation. For example, in retinal prostheses, the function of the photoreceptor in the retina that has lost its function is replaced by an image sensor, and the output of the image sensor is used as an electrical stimulus to higher-order neurons such as bipolar cells and retinal ganglion cells. Therefore, in a broad sense, visual prostheses can be treated as biomimetic devices. However, there are many difficulties in reproducing visual sensations that are close to natural vision. For example, ON-type cells that are excited by light irradiation and OFF-type cells that are suppressed by irradiation coexist in the actual retina, but electrical stimulation cannot distinguish these cells and stimulate them simultaneously. In addition, the spatial location of the phosphene, which is a light sensation caused by artificial methods such as electrical stimulation, is affected by eye movements, but an image sensor of retinal prostheses is usually fixed to the patient's head and cannot acquire information of eye movement. Therefore, even if the same retinal region is stimulated, the perceived location of phosphene may differ. Various studies have been performed to overcome these dissociations between natural vision and artificial vision.<br/>In this talk, we will first outline the principles of visual prostheses and the current state of research. Next, comparisons of the light sensation induced by visual prostheses with natural vision will be presented. Finally, I will introduce the latest research aimed at realizing more naturalistic phosphene in visual prostheses.