pulse2percept: A Python-based simulation framework for bionic vision

Abstract

By 2020 roughly 200 million people worldwide will suffer from photoreceptor diseases such as retinitis pigmentosa and age-related macular degeneration, and a variety of retinal sight restoration technologies are being developed to target these diseases. One technology, analogous to cochlear implants, uses a grid of electrodes to stimulate remaining retinal cells. Two brands of retinal prostheses are currently approved for implantation in patients with late stage photoreceptor disease. Clinical experience with these implants has made it apparent that the vision restored by these devices differs substantially from normal sight. To better understand the outcomes of this technology, we developed pulse2percept, an open-source Python implementation of a computational model that predicts the perceptual experience of retinal prosthesis patients across a wide range of implant configurations. A modular and extensible user interface exposes the different building blocks of the software, making it easy for users to simulate novel implants, stimuli, and retinal models. We hope that this library will contribute substantially to the field of medicine by providing a tool to accelerate the development of visual prostheses.

Simulation of visual perception and learning with a retinal prosthesis

“…However, different models would be required for different kinds of devices (e.g. see Beyeler et al 2017). Furthermore, different electrodes in the device could activate nearby cells differently, based on factors such as their physical apposition to the retina and local damage due to disease, factors not accounted for here but that could potentially be calibrated in the clinic.…”

“…Because the reconstruction framework produces images as predictions of visual experience, it could be used by clinicians to compare with and interpret patient verbal reports. Such feedback may help to fine-tune the assumptions and parameters of the simulation (see Beyeler et al 2017), such as the degree of ON dominance or learning (see above). Note that variations in performance that are relevant to the clinic, such as surgical approach, were not considered here.…”

“…Nonetheless, the degree to which technical improvements and learning can in principle improve outcomes remains poorly understood, and current understanding of the experience of artificial vision itself is limited. Thus, it would be valuable to have a computational framework to predict the visual experience that retinal prostheses provide to patients, as well as how that visual experience could potentially improve with learning (Beyeler et al 2017). Such a framework could help researchers and clinicians interpret what patients see and how they learn, and could guide future technical development by identifying factors that limit performance.…”

“…The reconstruction framework approach has the advantage of providing a direct assessment of the visual sensations potentially experienced by the patient, rather than an indirect representation in terms of patterns of electrical activity in the optic nerve. Unlike existing approaches (Beyeler et al 2017), it has a theoretical foundation, and in principle it can be extended from the starting point presented here to include all known aspects of retinal signals and retinal prostheses. Finally, the approach provides a straightforward way to reason about how a patient with a prosthesis could potentially learn to use it effectively, a fundamental aspect of neural interfaces for which few quantitative reasoning tools are available.…”

Biophysical model of axonal stimulation in epiretinal visual prostheses

“…Because epiretinal prostheses sit on top of the optic fiber layer, these devices may accidentally stimulate passing axon fibers, which could antidromically activate cell bodies located peripheral to the point of stimulation [3]- [5]. This can cause nontrivial perceptual distortions [6]- [8] that may severely limit the quality of the generated visual experience [9], [10].…”