Neural interfaces: implanted, wearable and focused ultrasound

How do implanted, skin-surface and ultrasound neural interfaces differ in surgery, demonstrated use, participant numbers and regulatory status?

Neural interfaces reach the nervous system by very different routes. Some implants sit inside the brain, in blood vessels or under the retina, while spinal stimulation and focused ultrasound in these sources work without implanted hardware. Each cell below is tied to its source and typed by confidence.

Neural interfaces: implanted, wearable and focused ultrasound. Each cell shows a value, how well it is evidenced, and its sources.
Option	Surgery neededWhether the approach needs surgery to place a device, and the route used. Yes or no, with the route stated.	Demonstrated result in peopleThe reported result in people, with the measure and its units (for example words per minute, percent or logMAR).	People treated or testedNumber of participants in the cited studies or trials, counted as reported (people), with the study year.	Regulatory statusMarketing approval or clearance stated in the sources, with the region and the indication. Trial authorisation is shown separately.
Implanted brain-computer interfaces	Intracortical arrays need brain surgery. The SWITCH Stentrode was delivered through the jugular vein, without open surgery.Demonstrated[1][2]	Attempted speech decoded into text at 62 words per minute, in one participant with ALS (lab study, 2023).Demonstrated[1]	1 participant in the 2023 speech study, 5 enrolled and 4 implanted in the SWITCH safety study (2023), and 6 in the COMMAND study (company-reported, 2024).Reported, unverified[1][2][3]	No marketing approval in these sources. COMMAND has an FDA investigational device exemption, which is trial authorisation only.Reported, unverified[3]
Spinal cord stimulation	No surgery reported. Stimulation is applied externally over the cervical spinal cord.Demonstrated[4]	72% of the 60 analysed people met responder criteria for both strength and function, in a single-arm trial with no control group.Demonstrated[4]	60 analysed participants with chronic tetraplegia, in a single-arm trial (2024).Demonstrated[4]	Company states FDA clearance in the US for incomplete cervical injury (C2 to C8), and CE marking in the EU. The page gives no dates.Reported, unverified[5]
Focused ultrasound	No open surgery. The device is noninvasive, and the high-intensity procedure is outpatient and needs no anaesthesia.Demonstrated[7]	In the pivotal trial, the treated group's CRST tremor score improved 46.9% from baseline at 3 months (95% CI 40.3% to 53.5%), in a sham-controlled trial of high-intensity treatment.Demonstrated[7]	76 in the pivotal tremor trial (56 treated, 20 sham). 690 participants across 34 low-intensity motor and sensory studies (2025 review).Demonstrated[7][6]	FDA-approved for high-intensity treatment of essential tremor not responding to medication, and for Parkinson disease in 2021. No low-intensity approval is reported in these sources.Reported, unverified[7][6]
Retinal prostheses	Yes. PRIMA is implanted under the retina in surgery, and 21 of its 26 serious adverse events occurred within two months of surgery.Demonstrated[9]	PRIMA: 26 of 32 assessed people (81%) had a clinically meaningful improvement in visual acuity at 12 months, in a single-group study in geographic atrophy.Demonstrated[9]	PRIMA: 38 implanted (2025). Argus II: 30 subjects in the 2012 interim trial, with best acuity 20/1260.Demonstrated[9][8]	Argus II: FDA approval in 2013 after humanitarian device exemption status, with production temporarily suspended in May 2019. PRIMA: approval not stated in sources.Reported, unverified[10][9]
Everyday BCIs goal	No reliable public figure	Attempted speech decoded into text in one participant (2023), in a research study.Demonstrated[1]	6 people in the COMMAND study, company-reported (2024). No larger controlled trial in these sources.Reported, unverified[3]	No reliable public figure

Neural interfaces: implanted, wearable and focused ultrasound. Each cell shows a value, how well it is evidenced, and its sources.

Option

Surgery neededWhether the approach needs surgery to place a device, and the route used. Yes or no, with the route stated.

Demonstrated result in peopleThe reported result in people, with the measure and its units (for example words per minute, percent or logMAR).

People treated or testedNumber of participants in the cited studies or trials, counted as reported (people), with the study year.

Regulatory statusMarketing approval or clearance stated in the sources, with the region and the indication. Trial authorisation is shown separately.

Implanted brain-computer interfaces

Intracortical arrays need brain surgery. The SWITCH Stentrode was delivered through the jugular vein, without open surgery.Demonstrated[1][2]

Attempted speech decoded into text at 62 words per minute, in one participant with ALS (lab study, 2023).Demonstrated[1]

1 participant in the 2023 speech study, 5 enrolled and 4 implanted in the SWITCH safety study (2023), and 6 in the COMMAND study (company-reported, 2024).Reported, unverified[1][2][3]

No marketing approval in these sources. COMMAND has an FDA investigational device exemption, which is trial authorisation only.Reported, unverified[3]

Spinal cord stimulation

No surgery reported. Stimulation is applied externally over the cervical spinal cord.Demonstrated[4]

72% of the 60 analysed people met responder criteria for both strength and function, in a single-arm trial with no control group.Demonstrated[4]

60 analysed participants with chronic tetraplegia, in a single-arm trial (2024).Demonstrated[4]

Company states FDA clearance in the US for incomplete cervical injury (C2 to C8), and CE marking in the EU. The page gives no dates.Reported, unverified[5]

Focused ultrasound

No open surgery. The device is noninvasive, and the high-intensity procedure is outpatient and needs no anaesthesia.Demonstrated[7]

In the pivotal trial, the treated group's CRST tremor score improved 46.9% from baseline at 3 months (95% CI 40.3% to 53.5%), in a sham-controlled trial of high-intensity treatment.Demonstrated[7]

76 in the pivotal tremor trial (56 treated, 20 sham). 690 participants across 34 low-intensity motor and sensory studies (2025 review).Demonstrated[7][6]

FDA-approved for high-intensity treatment of essential tremor not responding to medication, and for Parkinson disease in 2021. No low-intensity approval is reported in these sources.Reported, unverified[7][6]

Retinal prostheses

Yes. PRIMA is implanted under the retina in surgery, and 21 of its 26 serious adverse events occurred within two months of surgery.Demonstrated[9]

PRIMA: 26 of 32 assessed people (81%) had a clinically meaningful improvement in visual acuity at 12 months, in a single-group study in geographic atrophy.Demonstrated[9]

PRIMA: 38 implanted (2025). Argus II: 30 subjects in the 2012 interim trial, with best acuity 20/1260.Demonstrated[9][8]

Argus II: FDA approval in 2013 after humanitarian device exemption status, with production temporarily suspended in May 2019. PRIMA: approval not stated in sources.Reported, unverified[10][9]

Everyday BCIs goal

No reliable public figure

Attempted speech decoded into text in one participant (2023), in a research study.Demonstrated[1]

6 people in the COMMAND study, company-reported (2024). No larger controlled trial in these sources.Reported, unverified[3]

No reliable public figure

What the table shows

The intracortical arrays and the PRIMA retinal implant need surgery. The Stentrode, ARCEX and focused ultrasound do not need open surgery.
Implanted BCI rows report between one and six people, while the retinal, spinal and ultrasound rows report 30 or more people in their main studies.
The approvals in the table cover different uses: high-intensity ultrasound for essential tremor and Parkinson disease, Argus II for retinitis pigmentosa, and a company-stated clearance for incomplete spinal cord injury.
No implanted BCI in this table shows marketing approval. The COMMAND entry is trial authorisation only.

Read with care

Figures are not like for like: a 2012 interim trial, a 2023 single-person study, a 2024 company-reported trial and a 2025 single-group study measure different things.
Some figures come from company announcements or trade press (COMMAND, the ONWARD ARCEX page and Practical Neurology). These have not been checked against regulator databases.
The spinal stimulation trial had no control group or sham, so its gains cannot be fully separated from rehabilitation alone.
The 690 participants come from 34 separate studies, so they do not describe one consistent effect.

Sources

peer reviewed paper
A high-performance speech neuroprosthesis
Francis R. Willett, Erin M. Kunz, Chaofei Fan et al.. Nature, 23 Aug 2023. Accessed 21 Oct 2026.
peer reviewed paper
Assessment of Safety of a Fully Implanted Endovascular Brain-Computer Interface for Severe Paralysis in 4 Patients: The Stentrode With Thought-Controlled Digital Switch (SWITCH) Study
Peter Mitchell, Sarah C. M. Lee, Peter E. Yoo et al.. JAMA Neurology, 9 Jan 2023. Accessed 21 Oct 2026.
company announcement
Synchron Announces Positive Results from U.S. COMMAND Study of Endovascular Brain-Computer Interface
Synchron, Inc. (company announcement, reprinted on Neurofounders), 30 Sept 2024. Accessed 21 Oct 2026.
peer reviewed paper
Non-invasive spinal cord electrical stimulation for arm and hand function in chronic tetraplegia: a safety and efficacy trial
Chet Moritz, Grégoire Courtine. Nature Medicine, 20 May 2024. Accessed 21 Oct 2026.
company announcement
ARC-EX (ARCEX System) product information
ONWARD Medical, 21 Oct 2026. Accessed 21 Oct 2026.
review article
Neuromodulation effects of low-intensity transcranial focused ultrasound in human, a systematic review focusing on motor and sensory functions
Heng-Chen Ho, Wen-Shiang Chen, Ming-Yen Hsiao. Journal of NeuroEngineering and Rehabilitation, 28 Nov 2025. Accessed 21 Oct 2026.
news
The FDA Has Approved a Focused Ultrasound Device for Essential Tremor
Practical Neurology, 11 Jul 2025. Accessed 21 Oct 2026.
peer reviewed paper
Interim results from the international trial of Second Sight's visual prosthesis
Mark S. Humayun, Jessy D. Dorn, Lyndon Da Cruz et al.. Ophthalmology, Apr 2012. Accessed 21 Oct 2026.
peer reviewed paper
Subretinal Photovoltaic Implant to Restore Vision in Geographic Atrophy Due to AMD
Frank G. Holz, José Alain Sahel. New England Journal of Medicine, 20 Oct 2025. Accessed 21 Oct 2026.
other
Retinal Implants for RP: An Update on Argus II and Others
Rohan Bajaj, Gislin Dagnelie, James T. Handa et al.. American Academy of Ophthalmology (EyeNet), 2 Dec 2019. Accessed 21 Oct 2026.