The DOI matches its Crossref record; 100 of the 109 references it cites were found in a registry or an index; 37 of the 52 sources its claims cite were checked for support.
Towards a Smart Bionic Eye : AI-powered artificial vision for the treatment of incurable blindness
Open accessAbstract
Objective. How can we return a functional form of sight to people who are living with incurable blindness? Despite recent advances in the development of visual neuroprostheses, the quality of current prosthetic vision is still rudimentary and does not differ much across different device technologies. Approach. Rather than aiming to represent the visual scene as naturally as possible, a Smart Bionic Eye could provide visual augmentations through the means of artificial intelligence-based scene understanding, tailored to specific real-world tasks that are known to affect the quality of life of people who are blind, such as face recognition, outdoor navigation, and self-care. Main results. Complementary to existing research aiming to restore natural vision, we propose a patient-centered approach to incorporate deep learning-based visual augmentations into the next generation of devices. Significance. The ability of a visual prosthesis to support everyday tasks might make the difference between abandoned technology and a widely adopted next-generation neuroprosthetic device.
Overview
CoChat Summary
TL;DR
The paper proposes a “Smart Bionic Eye” that uses AI-based scene understanding to provide task-specific visual augmentations rather than attempting to restore natural vision. It argues that a patient-centered, iterative design process with virtual prototyping and real-world task modes could make next-generation visual prostheses more useful and more likely to be adopted.
Key findings
- Current prosthetic vision remains rudimentary and does not differ much across device technologies; users report distorted phosphenes that vary across subjects and electrodes and often fail to assemble into complex percepts.
- Three visual prostheses have been commercialized (Second Sight’s Argus II, Retina Implant AG’s Alpha-AMS, and Pixium Vision’s IRIS II), and only one regulatory-approved gene therapy exists (Luxturna), yet existing devices generally provide only improved localization of high-contrast objects and basic orientation & mobility.
- The authors propose AI-powered visual augmentations for tasks such as visual search, screen reading, conversation, and navigation, including visual question answering to answer questions like “Where did I put my keys again?”
- They note that Retina Implant AG dissolved in March 2019 and cited lack of “concrete benefit in everyday life” as a main reason, suggesting that supporting everyday tasks may determine whether a next-generation device is abandoned or widely adopted.
Methods
- This is a perspective/letter, not an empirical study; no new experimental data, participants, or statistical analysis are reported.
- It synthesizes literature on visual prostheses, deep learning–based computer vision, simulated prosthetic vision, and human-computer interaction/patient-centered design.
- It proposes an iterative workflow beginning and ending with the patient: identify end-user information needs, develop task-specific visual augmentations, refine them using qualitative feedback and behavioral performance, then test promising strategies in real prosthesis patients.
Limitations
- The paper states challenges and limitations including risk/benefit assessment versus less-invasive alternatives, the unresolved neural code of vision and perceptual distortions, robustness and safety of AI in real-world settings, and engineering constraints for low-power portable edge devices.
- Evident from the text: no empirical validation of the proposed Smart Bionic Eye or its AI augmentations is presented; most cited augmentation studies used hypothetical future devices, sighted subjects, or simplified simulations, and few tested real prosthesis patients.
- Evident from the text: the proposal depends on future next-generation hardware and AI maturity, and no clinical outcomes, adoption data, or patient preference results are reported.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
22 Valid · 4 Not valid · 6 Not checked · 12 Own results and interpretations
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“As pointed out by Htike et al. (2020) and Erickson-Davis and Korzybska (2021), the majority of research on visual prostheses (and more generally: low vision aids) has focused on the technical aspects rather than the usability of these devices.”
§ Towards a Smart Bionic Eye: AI-Powered Artificial Vision for the Treatment of Incurable Blindness > Towards a Smart Bionic Eye > The Proposed Solution
What the sources say
- Partly supportedcites Htike et al., 2020 sourceExperimental — accuracy not yet measured
The Results section, under “Human Computer Interaction Aspects” (page unavailable), says most studies of head-mounted displays as mobility aids did not mention usability-related HCI aspects. That supports the concern about neglected usability, but the review does not establish the same pattern for visual prostheses or low-vision aids generally, nor does this passage say those studies focused on technical aspects instead. (LLM, conf=0.86)
- Partly supportedcites Erickson-Davis and Korzybska, 2021 sourceExperimental — accuracy not yet measured
The Introduction says research on retinal prostheses has overlooked users’ subjective accounts, including what it is like to learn to use and live with the devices. The Discussion says qualitative reports have been neglected relative to functional outcome measures. The paper does not establish that a majority of research focuses on technical aspects, or extend that assessment to low-vision aids generally; page numbers are unavailable in our copy. (LLM, conf=0.84)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“new computer vision algorithms are needed to process the unconventional output of neuromorphic sensors to unlock their potential (Gallego et al., 2022; Sanchez-Garcia et al., 2022).”
§ Towards a Smart Bionic Eye: AI-Powered Artificial Vision for the Treatment of Incurable Blindness > Towards a Smart Bionic Eye > Challenges & Limitations
What the sources say
- Supportedcites Gallego et al., 2022 source
The Abstract says that novel methods are required to process event cameras’ unconventional output to unlock their potential. Section 2.3 further explains that their event output differs fundamentally from conventional images, so frame-based vision algorithms are not directly applicable. (LLM, conf=0.95)
- Could not verifycites Sanchez-Garcia et al., 2022 sourceExperimental — accuracy not yet measured
The Introduction (p. 1) discusses spiking neural networks as a route to efficient artificial vision, and the Abstract (p. 1) describes a model tested on Fashion MNIST stimuli. Searches for neuromorphic sensors, their unconventional output, and algorithms needed to process it found no passage addressing that claim. (LLM, conf=0.76)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“In addition, there often exists a numeric or symbolic forward model, constrained by empirical data, that can predict a neuronal or ideally perceptual response to the applied stimulus (Bosking et al., 2017; Beyeler et al., 2019).”
§ Towards a Smart Bionic Eye: AI-Powered Artificial Vision for the Treatment of Incurable Blindness > Towards a Smart Bionic Eye > Challenges & Limitations
What the sources say
- Supportedcites Bosking et al., 2017 source
The Model section gives equations that predict cortical activation from stimulation current and then predict phosphene size, a perceptual response. The Results and model-fitting passages show that the model was constrained by empirical data and compared with subjects’ reported phosphene sizes. (LLM, conf=0.92)
- Supportedcites Beyeler et al., 2019 source
The Discussion states that a computational model predicts percepts elicited by retinal implants using the spatial layout of ganglion axon pathways. The Methods describe fitting model parameters to subjects’ drawings and using them to predict held-out percept shapes, making this paper an example of an empirically constrained forward model. (LLM, conf=0.92)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“These AI systems can make serious mistakes that no sane human would make (Hole and Ahmad, 2021).”
§ Towards a Smart Bionic Eye: AI-Powered Artificial Vision for the Treatment of Incurable Blindness > Towards a Smart Bionic Eye > Challenges & Limitations
What the sources say
- Supportedcites Hole and Ahmad, 2021 source
Section 3.2, “Limitations of narrow AI,” states the claim almost verbatim: narrow AI systems can make serious mistakes no sane human would make. It gives misclassified traffic signs as an example. (LLM, conf=0.95)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“more recent work has highlighted that the percepts resulting from multi-electrode stimulation cannot be explained by a summative model based on single-electrode phosphenes (Beauchamp et al., 2020; Barry et al., 2020; Fernández et al., 2021).”
§ Towards a Smart Bionic Eye: AI-Powered Artificial Vision for the Treatment of Incurable Blindness > Introduction
What the sources say
- Supportedcites Beauchamp et al., 2020 source
The Results section, “Comparison of Dynamic Current Steering with other stimulation paradigms,” reports that simultaneous stimulation of five electrodes produced two large phosphenes rather than the five expected from their individual effects. This directly supports the claim that multi-electrode percepts are not explained by simply summing single-electrode phosphenes. (LLM, conf=0.93)
- Not matchedcites Barry et al., 2020
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- Partly supportedcites Fernández et al., 2021 sourceExperimental — accuracy not yet measured
The Discussion (pp. 10–11) says simultaneous stimulation often merged two phosphenes and sometimes produced perceptions that did not clearly correspond to electrode locations; characteristics of individual phosphenes could be lost. This supports the challenge to a simple summative account, but the paper does not establish that such a model cannot explain multi-electrode percepts generally: the Results (p. 6) also report cases in which two electrodes evoked two discrete phosphenes. (LLM, conf=0.84)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
17 more claims · 5 with their cited sources checked
Create a free account to see all 44 claims and every source verdict.
- It may state two separate claims as one. 16
- Probably not a substantial claim. 8
Citations
Cites
Microsoft COCO: Common Objects in Context
Lin, T.-Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Dollár, P., & Zitnick, C. L. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10602-1_48 · Cited by 45,460 · cited 1×
“For example, a user may verbally instruct the Smart Bionic Eye to locate misplaced keys, and the system would respond visually by segmenting the keys in the prosthetic image while the user is looking around the room (room image reprinted under CC-BY from Lin et al., 2014).”
BackgroundVQA: Visual Question Answering
Antol, S., Agrawal, A., Lu, J., Mitchell, M., Batra, D., Zitnick, C. L., & Parikh, D. (2015) · 2015 IEEE International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2015.279 · Cited by 4,487
Gallego, G., Delbruck, T., Orchard, G., Bartolozzi, C., Taba, B., Censi, A., Leutenegger, S., Davison, A. J., Conradt, J., Daniilidis, K., et al. (2022) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2020.3008413 · Cited by 2,566
Digging Into Self-Supervised Monocular Depth Estimation
Godard, C., Aodha, O. M., Firman, M., & Brostow, G. (2019) · 2019 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00393 · Cited by 2,561
Russell, S., Bennett, J., Wellman, J. A., Chung, D. C., Yu, Z.-F., Tillman, A., Wittes, J., Pappas, J., Elci, O., McCague, S., et al. (2017) · The Lancet · doi:10.1016/s0140-6736(17)31868-8 · Cited by 1,952 · cited 1×
“Although recent advances in gene and stem cell therapies (e.g. Russell et al 2017, da Cruz et al 2018; for a recent review see McGregor, 2019) as well as retinal sheet transplants (e.g. Foik et al 2018, Gasparini et al 2019; for a recent commentary see Beyeler 2019) are showing great promise as…”
Handbook of Usability Testing: How to Plan, Design, and Conduct Effective Tests
Rubin, J. N. (1994) · Cited by 1,740
da Cruz, L., Fynes, K., Georgiadis, O., Kerby, J., Luo, Y. H., Ahmado, A., Vernon, A., Daniels, J. T., Nommiste, B., Hasan, S. M., et al. (2018) · Nature Biotechnology · doi:10.1038/nbt.4114 · Cited by 706 · cited 1×
“Although recent advances in gene and stem cell therapies (e.g., Russell et al., 2017, da Cruz et al., 2018; for a recent review see McGregor, 2019) as well as retinal sheet transplants (e.g., Foik et al., 2018, Gasparini et al., 2019; for a recent commentary see Beyeler, 2019) are showing great…”
BackgroundGallego, G., Delbrück, T., Orchard, G., Bartolozzi, C., Taba, B., Censi, A., Leutenegger, S., Davison, A. J., Conradt, J., Daniilidis, K., et al. (2020) · Zurich Open Repository and Archive (University of Zurich) · doi:10.3929/ethz-b-000521357 · Cited by 591
Artificial Vision for the Blind by Connecting a Television Camera to the Visual Cortex
Dobelle, W. H. (2000) · ASAIO Journal · doi:10.1097/00002480-200001000-00002 · Cited by 584 · cited 2×
“…described as ‘a star in the sky’) in response to single-electrode stimulation of the visual cortex (Dobelle and Mladejovsky 1974, Evans et al 1979, Dobelle 2000, Bosking et al 2017), more recent work has highlighted that the percepts resulting from multi-electrode stimulation cannot be…”
Rizzo, J. F., Wyatt, J., Loewenstein, J., Kelly, S., & Shire, D. (2003) · Investigative Opthalmology & Visual Science · doi:10.1167/iovs.02-0817 · Cited by 478 · cited 1×
“In the case of epiretinal implants, these distortions are largely due to inadvertent activation of passing axon fibers (Rizzo et al 2003, Beyeler et al 2019), but other device technologies based on electrical stimulation of visual cortex or optogenetics may face related issues.”
Research works printed but not linked: 1
- Barry, 2020, Video-mode percepts are smaller than sums of single-electrode phosphenes with the Orion® visual cortical prosthesis, Investigative Ophthalmol. Vis. Sci., 61, 927
Cited by
Neural interfaces: Bridging the brain to the world beyond healthcare
Xu, S., Liu, Y., Lee, H., & Li, W. (2024) · Exploration · doi:10.1002/exp.20230146 · Cited by 106
A systematic review of extended reality (XR) for understanding and augmenting vision loss
Kasowski, J., Johnson, B. A., Neydavood, R., Akkaraju, A., & Beyeler, M. (2023) · Journal of Vision · doi:10.1167/jov.23.5.5 · Cited by 62
“By design, these tasks require a more immersive setup that allows for the incorporation of head and eye movements as well as locomotion (Kasowski & Beyeler, 2022).”
Self-citationvan der Grinten, M., de Ruyter van Steveninck, J., Lozano, A., Pijnacker, L., Rueckauer, B., Roelfsema, P., van Gerven, M., van Wezel, R., Güçlü, U., & Güçlütürk, Y. (2024) · eLife · doi:10.7554/elife.85812 · Cited by 39
Flexible and stretchable synaptic devices for wearable neuromorphic electronics
Lee, H.-S., Ro, J.-S., Ko, G.-M., & Park, H. (2023) · Flexible and Printed Electronics · doi:10.1088/2058-8585/ad0a37 · Cited by 24
Assistive technology use in domestic activities by people who are blind
Turkstra, L. M., Bhatia, T., Os, A. V., & Beyeler, M. (2025) · Scientific Reports · doi:10.1038/s41598-025-91755-w · Cited by 18
Self-citation
Metadata
- Authors
- Beyeler, M., & Sanchez-Garcia, M.
- Affiliations
- University of California, Santa Barbara (US)
- Venue
- Journal of Neural Engineering
- Publisher
- IOP Publishing
- Published
- 2022-12-01
- Volume
- 19
- Issue
- 6
- Pages
- 063001
- ISSN
- 1741-2552 (ISSN-L)
- 1741-2560
- PMC
- PMC10507809
- OpenAlex
- W4310123661
- Cited by
- 35
- References
- 109
- Citations by year
- 2026: 6 · 2025: 15 · 2024: 7 · 2023: 6 · 2022: 1
- FWCI
- 2.50
- Type
- journal article
- Language
- en
- Open access
- green
- doi.org
- Licence
- cc-by
- publishingsupport.iopscience.iop.org/iop-standard/v1 (vor, 6 d)
- iopscience.iop.org/info/page/text-and-data-mining (tdm, 6 d)
- Locations
- Journal of Neural Engineering (publishedVersion)
- OSF Preprints (OSF Preprints) (submittedVersion, cc-by)
- PubMed (publishedVersion)
- PubMed Central (acceptedVersion, cc-by)
- eScholarship (California Digital Library) (submittedVersion)
- OSF Preprints (OSF Preprints) (submittedVersion)
- Funders
- U.S. National Library of Medicine (DP2LM014268)
- Topics
- Life Sciences › Neuroscience › Cellular and Molecular Neuroscience › Neuroscience and Neural Engineering
- Life Sciences › Neuroscience › Cognitive Neuroscience › Tactile and Sensory Interactions
- Life Sciences › Neuroscience › Cognitive Neuroscience › EEG and Brain-Computer Interfaces
- Keywords
- retinal prosthesis, visual prosthesis, artificial vision, blindness, visual augmentation, deep learning, scene understanding, face recognition (cognition), outdoor navigation, self-care, patient-centered design, neuroprosthetics
- MeSH terms
- Facial Recognition, Artificial Intelligence, Blindness, Humans, Quality of Life, Visual Prosthesis
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Towards a Smart Bionic Eye : AI-powered artificial vision for the treatment of incurable blindness. https://app.cochat.ai/papers/10.1088/1741-2552/aca69d
- The paper
- Beyeler, M., & Sanchez-Garcia, M. (2022). Towards a Smart Bionic Eye : AI-powered artificial vision for the treatment of incurable blindness. Journal of Neural Engineering. https://doi.org/10.1088/1741-2552/aca69d
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.