CoChat
Sign Up
Cited by 17Neuroscience and Neural EngineeringEEG and Brain-Computer InterfacesTactile and Sensory Interactions

Toward a personalized closed-loop stimulation of the visual cortex: Advances and challenges

Open access
Grani, Fabrizio, Soto-Sánchez, Cristina, Fimia, Antonio, Fernández, Eduardo·2022·Frontiers in Cellular Neuroscience
DOI 10.3389/fncel.2022.1034270PMC PMC9792612
Open this paper in CoChat — free accountOpen PDF

Abstract

Current cortical visual prosthesis approaches are primarily unidirectional and do not consider the feed-back circuits that exist in just about every part of the nervous system. Herein, we provide a brief overview of some recent developments for better controlling brain stimulation and present preliminary human data indicating that closed-loop strategies could considerably enhance the effectiveness, safety, and long-term stability of visual cortex stimulation. We propose that the development of improved closed-loop strategies may help to enhance our capacity to communicate with the brain.

AI snapshot

Written from the paper’s full text.

TL;DR

Current cortical visual prostheses are mostly open-loop, but closed-loop stimulation using neural recordings could improve effectiveness, safety, and long-term stability. The authors review recent closed-loop approaches and present preliminary human intracortical data supporting automatic perception-threshold detection and LFP-phase-locked stimulation.

Key findings

  • Closed-loop strategies are already used or studied in epilepsy, obsessive-compulsive disorder, depression, Parkinson’s disease, cochlear implants, spinal cord stimulation, and retinal prostheses; the paper catalogs these in Table 1.
  • In ongoing human intracortical experiments, neuronal activity after stimulation increased when stimulation intensity was enough to induce perception, shown at 40 μA; neural-signal-based psychometric curves for a single electrode using 0–140 μA were practically indistinguishable from standard user-feedback curves.
  • A real-time LFP phase-estimation algorithm tested on intracortical signals from a human blind volunteer’s visual cortex at 30 kHz achieved an error of ±20° in phase estimation.
  • The authors estimate that at least 625 electrodes implanted in visual areas would be needed for reading at lower speeds and navigating complex visual environments.

Methods

  • Narrative review/perspective synthesizing recent developments in closed-loop brain stimulation and cortical visual prostheses, not a systematic review.
  • Includes preliminary human data from ongoing experiments with intracortical microelectrodes in blind volunteers; studies were reviewed and approved by Hospital General Universitario de Elche Clinical Research Committee.
  • Main analysis is qualitative: a table of closed-loop neural stimulation examples, proposed flowcharts for automatic perception-threshold adjustment and LFP-phase-locked stimulation, and comparison of neural-signal versus user-feedback psychometric curves.

Limitations

  • Stated by the paper: signal stability over time, influence of movements on signal quality, higher energy consumption for closed-loop stimulation, stimulation artifacts corrupting recordings, uncertainty about whether the same LFP phase is valid for all electrodes, phase-estimation uncertainty, and phase-locked stimulation limiting time resolution to roughly every 100 ms because ongoing oscillations are around 10 Hz.
  • Stated by the paper: results must be confirmed in real-life environments, and there is still insufficient information about the LFP period corresponding to maximum excitability and how many feedback channels can be reliably provided in parallel.
  • Evident from the text: the human data are preliminary, from ongoing experiments with limited sample detail, and the proposed algorithms are not yet clinically validated.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3389/fncel.2022.1034270

Cites

65 of 65 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 46 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    This approach can also be used to construct psychometric curves (relation between stimulus intensity and perception probability) that are practically indistinguishable of the standard psychometric curves using users' feed-back.

    “This approach can also be used to construct psychometric curves (relation between stimulus intensity and perception probability) that are practically indistinguishable of the standard psychometric curves using users’ feed-back.”
    Quote in the paperSubstantial claim
  • Findingown data

    In ongoing experiments with intracortical microelectrodes in blind volunteers, the neuronal activity after stimulation increases when the stimulation intensity is enough to induce perception (40 μA in this case).

    “Figure 1B shows an example from our ongoing experiments with intracortical microelectrodes in blind volunteers in which the neuronal activity after stimulation increases when the stimulation intensity is enough to induce perception (40 μA in this case).”
    Quote in the paperSubstantial claim
  • Premisecitation

    There is no treatment for all causes of blindness.

    “unfortunately, there is no treatment for all causes of blindness (Fernandez, 2018; Fernandez et al., 2020).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Fernandez, 2018 source

      Not checked in the free run.

    • cites Fernandez et al., 2020 source

      Not checked in the free run.

  • Premisecitation

    The automatic tuning of stimulation thresholds in cochlear implants can be done by measuring the contraction of the stapedius muscle.

    “Thus, the automatic tuning of stimulation thresholds in cochlear implants can be done by measuring the contraction of the stapedius muscle.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Weiss et al., 2021 source

      Not checked in the free run.

  • Premisecitation

    A recent study in monkey visual cortex shows that the activity of neurons in V4 provides direct insight into the efficacy of the stimulation in V1.

    “In this framework, a recent study in monkey visual cortex shows that the activity of neurons in V4 provides direct insight into the efficacy of the stimulation in V1 (Chen et al., 2020).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Chen et al., 2020 source

      Not checked in the free run.

  • May combine two claims. 11
  • Classified as probably not a substantial claim. 16

  • May combine two claims.

    Findingcitation

    We have recently tested this algorithm using intracortical signals from the visual cortex of a human blind volunteer at a sampling rate of 30 kHz and we obtained an error of ±20° in the LFP phase estimation.

    “We have recently tested this algorithm using intracortical signals from the visual cortex of a human blind volunteer at a sampling rate of 30 kHz and we obtained an error of ±20° in the LFP phase estimation (Grani et al., 2022b).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Grani et al., 2022b

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    The biological and engineering problems for the success of cortical implants are much more complex than originally believed and involve long-term biocompatibility issues and challenges related to encoding of visual information and delivery of information to implants.

    “However, the biological and engineering problems for the success of cortical implants are much more complex than originally believed and involve, for example, long-term biocompatibility issues and challenges related to the encoding of visual information and the delivery of information to implants (Fernandez et al., 2020).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Fernandez et al., 2020 source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    In humans, there is a causal relationship between cortical excitation and phosphenes perception so that the phase of pre-stimulus oscillatory activity seems linked to visual perception.

    “Moreover, it has been shown in humans that there is a causal relationship between cortical excitation and phosphenes perception so that the phase of pre-stimulus oscillatory activity seems linked to visual perception (Dugue et al., 2011)”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Dugue et al., 2011 source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Currently, bidirectional electrodes that allow stimulation and recording of neural activity at the same time exist, but are limited by the artifacts generated in the recordings by the stimulation.

    “Currently, bidirectional electrodes that allow stimulation and recording of neural activity at the same time exist, but are limited by the artifacts generated in the recordings by the stimulation (Xu et al., 2018).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Xu et al., 2018

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The movement output in spinal cord stimulation for motor recovery can be controlled in closed-loop changing the stimulation parameters.

    “while the movement output in spinal cord stimulation for motor recovery can be controlled in closed-loop changing the stimulation parameters (Wenger et al., 2014).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Wenger et al., 2014 source

      Not checked in the free run.

  • May combine two claims.

    Premisenone given

    Stimulating electrodes allow control of the dynamics of populations of neurons, but they do not provide insights into the electrophysiological activity of the neurons surrounding the electrodes.

    “Although stimulating electrodes allow control of the dynamics of populations of neurons, they do not provide insights into the electrophysiological activity of the neurons surrounding the electrodes.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The electrical stimulation of the spinal cord for pain therapy can be adjusted based on the evoked compound action potential (ECAP).

    “Also, the electrical stimulation of the spinal cord for pain therapy can be adjusted based on the evoked compound action potential (ECAP) (Mekhail et al., 2020)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Mekhail et al., 2020 source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Other measures like ECAP, EABR, and CAEPs have been studied to automatically fit cochlear implants, but none of them reached a clinical application.

    “Other measures like ECAP (McKay et al., 2013), electrically evocated auditory brainstem response (EABR) (Guenser et al., 2015), cortical auditory evoked potentials (CAEPs) (Visram et al., 2015) have been studied to automatically fit cochlear implants, but none of them reached a clinical application.”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites McKay et al., 2013 source

      Not checked in the free run.

    • cites Guenser et al., 2015 source

      Not checked in the free run.

    • cites Visram et al., 2015 source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Interpretationcitation

    The brain signals collected by the electrodes in the visual cortex could reveal a brain state in which it is easier to induce perception.

    “Furthermore, besides adjusting the thresholds, the brain signals collected by the electrodes in the visual cortex could reveal a brain state in which it is easier to induce perception.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Premisenone given

    Some recent works have led to the development of novel forms of neuromodulation, which are facilitating the ability to manipulate populations of neurons in near real-time.

    “Some recent works have led to the development of novel forms of neuromodulation, which are facilitating the ability to manipulate populations of neurons in near real-time.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Closed-loop procedures allow an improved control of some neurological conditions such as epilepsy, and can also be used for better control of obsessive-compulsive disorders and depression.

    “This procedures allow an improved control of some neurological conditions such as epilepsy (Ranjandish and Schmid, 2020; Farkhondeh Tale Navi et al., 2022), and can also be used for better control of obsessive-compulsive disorders and depression (Figee et al., 2022; Visser-Vandewalle et al., 2022).”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Ranjandish and Schmid, 2020

      Not checked in the free run.

    • cites Farkhondeh Tale Navi et al., 2022 source

      Not checked in the free run.

    • cites Figee et al., 2022 source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Closed-loop circuits exist in just about every part of the nervous system.

    “In this framework, we should consider that closed-loop circuits exist in just about every part of the nervous system (Farkhondeh Tale Navi et al., 2022; Khodagholy et al., 2022).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Farkhondeh Tale Navi et al., 2022 source

      Not checked in the free run.

    • cites Khodagholy et al., 2022 source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    Many questions need to be solved before closed-loop stimulation can be implemented in clinical devices able to continuously record and stimulate from hundreds of electrodes.

    “Although closed-loop stimulation might increase the safety, performance and usability of cortical visual prostheses, many questions need to be solved before it can be implemented in clinical devices able to continuously record and stimulate from hundreds of electrodes.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Perception can be inferred from EEG signals in the occipital cortex.

    “perception could be inferred from EEG signals in the occipital cortex (Gail et al., 2004)”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Gail et al., 2004 source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Interpretationreasoning

    The next frontier in cortical visual prosthesis may be the development of bidirectional implantable systems with enhanced abilities to modulate and manipulate populations of neurons in real-time.

    “As a result, the next frontier in cortical visual prosthesis may be the development of bidirectional implantable systems with enhanced abilities to modulate and manipulate populations of neurons in real-time.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    It has been estimated that we need at least 625 electrodes implanted in visual areas for reading (although at lower speeds) and navigating through complex visual environments.

    “It has been estimated that we need at least 625 electrodes implanted in visual areas for reading (although at lower speeds) and navigating through complex visual environments (Cha et al., 1992).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Cha et al., 1992 source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S4 (based). May combine two claims.

    Premisenone given

    There are still important issues and challenges to overcome, including the stability of signals over time, the influence of movements on signal quality, and the higher energy consumption needed to perform closed-loop stimulation.

    “However, there are still a number of important issues and challenges to overcome. For example, the stability of signals over time, the influence of movements on signal quality, and the higher energy consumption needed to perform closed-loop stimulation.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    LFP phase-dependent response to stimulation has been reported in experiments in rats and using non-invasive transcranial magnetic stimulation (TMS) in sighted humans.

    “This has been reported in experiments in rats (Allison-Walker et al., 2020) and using non-invasive transcranial magnetic stimulation (TMS) in sighted humans (Dugue et al., 2011), but the feasibility of LFP phase-locked stimulation with intracortical electrodes in humans still remains unexplored.”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Allison-Walker et al., 2020 source

      Not checked in the free run.

    • cites Dugue et al., 2011 source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Different algorithms have been proposed for continuous phase estimation in real-time.

    “although different algorithms have been proposed for continuous phase estimation in real-time (Kim et al., 2016; Blackwood et al., 2018)”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Kim et al., 2016 source

      Not checked in the free run.

    • cites Blackwood et al., 2018 source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S4 (based).

    Interpretationreasoning

    The neural population around each electrode could determine the preferred LFP phase.

    “Thus, the neural population around each electrode could determine the preferred LFP phase”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Interpretationnone given

    Adding a real-time brain signal analysis processor to a cortical visual prosthesis would increase the device's energy consumption.

    “adding a real-time brain signal analysis processor to the device would increase the energy consumption.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The sampling frequency often used to obtain reliable neural signals is 30 kHz.

  • Probably not a substantial claim: may not pass S4 (based).

    Interpretationnone given

    Adding a standard EEG cap to the prosthesis will decrease the overall wearability.

    “adding a standard EEG cap to the prosthesis will decrease the overall wearability”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

    Interpretationnone given

    Research on real-time closed-loop neural systems may prove essential for future cortical visual prostheses, especially when high-number microelectrodes are used.

    “Research on real-time closed-loop neural systems has built upon contributions from neuroscientists, engineers and clinicians, and may prove essential for future cortical visual prostheses, especially when high-number microelectrodes are used.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based). May combine two claims.

    Interpretationnone given

    Closed-loop approaches could be able to use information from neural recordings to adjust the optimal stimulation, reduce charge requirements, and improve stimulation performance.

    “These closed-loop approaches could be able to use information from neural recordings to adjust the optimal stimulation, reduce charge requirements, and improve stimulation performance.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning

    Bigger ΔI values will speed up the threshold finding at the cost of reducing the precision of the threshold.

    “Bigger ΔI values will speed up the threshold finding at the cost of reducing the precision of the threshold.”
    Quote in the paperNot a substantial claim
25 of 25

  • The paper relies on this without support.

    Current cortical visual prosthesis approaches are primarily unidirectional and do not incorporate any adaptive system for the modulation of the electrical stimulation used to induce visual perception.

    “Current cortical visual prosthesis approaches are primarily unidirectional and do not consider the feed-back circuits that exist in just about every part of the nervous system.”
    “However, current cortical visual prosthesis approaches are primarily unidirectional and do not incorporate any adaptive system for the modulation of the electrical stimulation used to induce visual perception.”

  • The paper relies on this without support.

    Patients with severe retinal degeneration, glaucoma, or optic atrophy cannot get benefit from a retinal prosthesis.

    “but patients with severe retinal degeneration, glaucoma, or optic atrophy cannot get benefit from a retinal prosthesis.”

  • The paper relies on this without support.

    The brain signals surrounding the electrodes contain information about the spread of currents and brain dynamics.

    “As the brain signals surrounding the electrodes contain information about the spread of currents and brain dynamics, we could determine if the currents used are enough to induce perception simply by measuring the brain response to electrical stimulation.”

  • The paper relies on this without support.

    For this to be possible, the brain signals during (or after) stimulation should have distinguishable features in case of perception or no perception.

    “For this to be possible, the brain signals during (or after) stimulation should have distinguishable features in case of perception or no perception.”

  • The paper relies on this without support.

    Perception thresholds could vary over time.

    “Moreover, perception thresholds could vary over time, requiring the users to calibrate each electrode many times.”

  • The paper relies on this without support.

    Finding the lowest current thresholds able to induce visual perceptions from each single electrode is a time-consuming procedure that requires the user’s feedback.

    “However, finding the lowest current thresholds able to induce visual perceptions from each single electrode is a time-consuming procedure that requires the user’s feedback.”

  • The paper relies on this without support.

    There is a need to reduce power consumption and the charge required to elicit reliable phosphenes.

    “On the other hand, there is a need to reduce power consumption and the charge required to elicit reliable phosphenes.”

  • The paper relies on this without support.

    The sampling frequency often used to obtain reliable neural signals is 30 kHz.

    “the sampling frequency often used to get reliable neural signals is 30 kHz.”

  • The paper relies on this without support.

    Signals collected during electrical stimulation are usually corrupted by stimulation artifacts.

    “the signals collected during electrical stimulation are usually corrupted by the stimulation artifacts.”

  • The authors state this as an assumption.

    Artifact features increase linearly with current intensity.

    “assuming that the artifact features increase linearly with the current intensity”

  • The authors state this as an assumption.

    Features linked to perception start to increase only after the perception threshold.

    “while features linked to perception should have a different behavior, starting to increase only after the perception threshold.”
11 of 11

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • 2020 reference lists an earlier journal volume than 2019 reference

    Minor

    The references list Frontiers in Neuroscience as volume 13 in 2019 but volume 11 in 2020. The same list gives volume 12 for 2018 and volume 16 for 2022, making the 2020 volume number inconsistent with its year.

1 of 1

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Grani, F., Soto-Sánchez, C., Fimia, A., & Fernández, E.
Affiliations
Universitat de Miguel Hernández d'Elx (ES)
Corresponding author
Fernández, Eduardo
ORCID
Venue
Frontiers in Cellular Neuroscience
Publisher
Frontiers Media SA
Published
2022-12-13
Volume
16
Pages
1034270-1034270
ISSN
1662-5102 (ISSN-L)
DOI
10.3389/fncel.2022.1034270
PMC
PMC9792612
OpenAlex
W4311354142
Semantic Scholar
eff3ebc6de8eac6637cef0a3a7e3e110cd460e7d
Cited by
16
References
65
Citations by year
2026: 4 · 2025: 8 · 2024: 1 · 2023: 3
FWCI
1.33
Type
journal article
Language
en
Open access
Licence
Locations
  • Frontiers in Cellular Neuroscience (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • European Commission (899287, 861423, HORIZON2020)
  • Generalitat Valenciana (PROMETEO/2019/119)
  • Ministerio de Ciencia e Innovación (PDC2022)
Topics
  • Life Sciences › Neuroscience › Cellular and Molecular Neuroscience › Neuroscience and Neural Engineering
  • Life Sciences › Neuroscience › Cognitive Neuroscience › EEG and Brain-Computer Interfaces
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Tactile and Sensory Interactions
Keywords
closed-loop stimulation, visual cortex stimulation, cortical visual prosthesis, brain-computer interface, feedback circuits, personalized stimulation

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.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-08.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-08.

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