A 2026 research demonstration used signals from one brain implant to decode attempted speech and upper-body gestures at the same time, letting two people with paralysis control personalized virtual avatars. In the interface, gestures animated the avatar while decoded speech appeared as text. The study shows a promising proof of concept—not a device people can buy or a service shown to work independently in daily life.
How the brain-controlled avatar worked
Researchers recorded neural activity from a single high-density electrocorticography (ECoG) implant. They built parallel decoders for attempted speech and upper-body gestures, then used those outputs to control personalized full-body virtual avatars in real time.
| # | Preview | Product | Price | |
|---|---|---|---|---|
| 1 |
|
Brain-Computer Interface Technologies: Accelerating Neuro-Technology for Human Benefit | $93.01 | Buy on Amazon |
| 2 |
|
Brain Computer Interface | $47.24 | Buy on Amazon |
| 3 |
|
Brain Computer Interface | $2.99 | Buy on Amazon |
| 4 |
|
Brain-Computer Interfacing: An Introduction | $54.27 | Buy on Amazon |
As an Amazon Associate I earn from qualifying purchases.
The team first tested whether signals from one cortical implant could decode isolated upper-limb and orofacial movements in three participants. The avatar demonstrations involved two participants. They attempted gestures, speech phrases, or both; decoded gestures animated the avatar, and decoded speech was displayed as text on screen. Nature Neuroscience study Avatar interface description
Two types of task
The researchers evaluated a cued copy task and a conversational-prompt task. They also described a qualitative virtual conversation: the participant’s brain-controlled avatar interacted with a second avatar operated by an experimenter entering speech and gesture commands on a keyboard. That interaction was a demonstration under research conditions, not an unsupervised communication service.
#1 Best Overall
What accuracy did the study report?
Results varied by participant and task. The figures below are individual experimental results, not estimates of how accurately the system would work for future users.
| Participant and task | Reported result | Reported chance level |
|---|---|---|
| Bravo-6, conversation paradigm, average across five blocks | Gesture decoding: 85.0% (95% confidence interval 70.0–100.0); speech decoding: 75.0% (95% confidence interval 55.0–90.0) | 9.1% for each decoder |
| Bravo-1r, median across three conversational blocks | Speech decoding: 100.0%; gesture decoding: 100.0% | 16.7% for speech; 20.0% for gestures |
These are small-sample task scores, not a general success rate. The authors also report that training with isolated and simultaneous data improved performance across behavioral contexts. Study accuracy results Study methods and findings
Rank #2
What the demonstration does—and does not—establish
The central advance is that signals from one cortical implant supported parallel control of speech and gesture, rather than requiring the avatar to handle only one output at a time. That could make a digital interaction feel more expressive: a user can convey words while also using body movement.
But the evidence is an early research demonstration with two participants in the avatar work. It does not establish that the system is ready for routine clinical use, works reliably in unsupervised everyday settings, or is available for purchase. The authors and institutional summary say that more participants and larger word and gesture sets are needed before practical use. Institutional summary and limitations
Rank #3
When the study took place
The trial began in November 2018. Data collection for this study ran from September 2022 through November 2025, and the paper was published in Nature Neuroscience on 14 September 2026. The publication date is not the same as the enrollment or data-collection period. Study dates and publication information
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
