Brain-Computer Interface Innovations

This 17 year old used AI to build a mind-controlled prosthetic for just $300. - A device that normally costs $450,000!!! Let that sink in for a moment... How is it that a high school student just built something that costs 1,500 times less than industry alternatives? What does that say about the system? About innovation? About who actually gets access to life changing technology? Benjamin Choi is just 17 years old. Using tiny electrodes on the forehead, his AI-driven prosthetic picks up brain activity and translates it into movement... and it doesn’t even require surgery or brain implants. He trained the AI on thousands of brainwave data points, wrote 23,000+ lines of code, and studied nearly 900 pages of calculus to make this happen. And it works! Yet, similar prosthetics cost around $450K. Why? Of course, high-end prosthetics are expensive for valid reasons: materials, research, clinical testing, regulatory approval, and customization for individual patients. Those costs aren’t just random numbers. But does that fully justify the price? How much of it is necessary cost, and how much is just the way the system works? This proves that low-cost, effective alternatives are possible if we rethink how we approach accessibility. If a teenager can match the performance of top-tier prosthetics for a fraction of the cost, why aren’t these solutions available to those who need them most? We can build the most incredible technology in the world, but if it's not accessible and affordable, have we really accomplished anything?

UC Berkeley and UCSF just brought real-time speech back to someone who couldn’t speak for 18 years (insane!). For people with paralysis and anarthria, the delay and effort of current AAC tools can make natural conversation nearly impossible. 𝗧𝗵𝗶𝘀 𝗻𝗲𝘄 𝗔𝗜-𝗱𝗿𝗶𝘃𝗲𝗻 𝗻𝗲𝘂𝗿𝗼𝗽𝗿𝗼𝘀𝘁𝗵𝗲𝘀𝗶𝘀 𝘀𝘁𝗿𝗲𝗮𝗺𝘀 𝗳𝗹𝘂𝗲𝗻𝘁, 𝗽𝗲𝗿𝘀𝗼𝗻𝗮𝗹𝗶𝘇𝗲𝗱 𝘀𝗽𝗲𝗲𝗰𝗵 𝗱𝗶𝗿𝗲𝗰𝘁𝗹𝘆 𝗳𝗿𝗼𝗺 𝗯𝗿𝗮𝗶𝗻 𝘀𝗶𝗴𝗻𝗮𝗹𝘀 𝗶𝗻 𝗿𝗲𝗮𝗹 𝘁𝗶𝗺𝗲 𝘄𝗶𝘁𝗵 𝗻𝗼 𝘃𝗼𝗰𝗮𝗹𝗶𝘇𝗮𝘁𝗶𝗼𝗻 𝗿𝗲𝗾𝘂𝗶𝗿𝗲𝗱. 1. Restored speech in a participant using 253-channel ECoG, 18 years after brainstem stroke and complete speech loss. 2. Trained deep learning decoders to synthesize audio and text every 80 ms based on silent speech attempts, with no vocal sound needed. 3. Streamed speech at 47.5 words per minute with just 1.12s latency = 8× faster than prior state-of-the-art neuroprostheses. 4. Matched the participant’s original voice using a pre-injury recording, bringing back not just words but vocal identity. Bimodal decoder architecture they used was cool. It's interesting how they got to low-latency and a synchronized output from the system. This was done by sharing a neural encoder and employing separate joiners and language models for both acoustic-speech units and text Other tidbits used was convolutional layers with unidirectional GRUs and LSTM-based language models. Absolutely love seeing AI used in practical ways to bring back joy and hope to people who are paralyzed!! Here's the awesome work: https://lnkd.in/ghqX5EB2 Congrats to Kaylo Littlejohn, Cheol Jun Cho, Jessie Liu, Edward Chang, Gopala Krishna Anumanchipalli, and co! I post my takes on the latest developments in health AI – 𝗰𝗼𝗻𝗻𝗲𝗰𝘁 𝘄𝗶𝘁𝗵 𝗺𝗲 𝘁𝗼 𝘀𝘁𝗮𝘆 𝘂𝗽𝗱𝗮𝘁𝗲𝗱! Also, check out my health AI blog here: https://lnkd.in/g3nrQFxW

Enhanced Brain Implant Translates Stroke Survivor’s Thoughts Into Nearly Instant Speech Using Artificial Intelligence The system harnesses technology similar to that of devices like Alexa and Siri, according to the researchers, and improves on a previous model Researchers connect stroke survivor Ann Johnson's brain implant to the experimental computer, which will allow her to speak by thinking words. Noah Berger A brain implant that converts neuron activity into audible words has given a stroke survivor with severe paralysis almost instantaneous speech. Ann Johnson became paralyzed and lost the ability to speak after suffering a stroke in 2005, when she was 30 years old. Eighteen years later, she consented to being surgically fitted with an experimental, thin, brain-reading implant that connects to a computer, officially called a brain-computer interface (BCI). Researchers placed the implant on her motor cortex, the part of the brain that controls physical movement, and it tracked her brain waves as she thought the words she wanted to say. As detailed in a study published Monday in the journal Nature Neuroscience, researchers used advances in artificial intelligence (A.I.) to improve the device’s ability to quickly translate that brain activity into synthetic speech—now, it’s almost instantaneous. The technology “brings the same rapid speech decoding capacity of devices like Alexa and Siri to neuroprostheses,” study co-author Gopala Anumanchipalli, a computer scientist at the University of California, Berkeley, says in a statement. Neuroprostheses are devices that can aid or replace lost bodily functions by connecting to the nervous system. “Using a similar type of algorithm, we found that we could decode neural data and, for the first time, enable near-synchronous voice streaming,” he adds. “The result is more naturalistic, fluent speech synthesis.” #AI #medicine #BrainComputerInterface #brainimplant #strokesurvivor #brainvoicesynthesis

This is the moment AI gave someone their voice back. It’s not science fiction anymore. For 18 years Ann has been paralysed and locked in. A stroke took her ability to speak but the neural signals remained. This video shows a historic breakthrough in brain computer interface technology. An electrocorticography grid decodes signals sent to her facial muscles. The AI translates them into speech on a digital avatar in real time. She says I think you are wonderful. Those are her first words spoken through an avatar using just her brain. This is where neuroscience meets artificial intelligence. We are moving beyond generative AI into restorative AI. It is about rebuilding the human connections we thought were lost forever. If AI can restore a lost voice, what other human capabilities could we rebuild next? #AI #HealthTech #Neuroscience #Innovation

𝗘𝘃𝗲𝗿𝘆 𝗰𝗵𝗶𝗹𝗱 𝗱𝗲𝘀𝗲𝗿𝘃𝗲𝘀 𝘁𝗼 𝗵𝗲𝗮𝗿 𝘁𝗵𝗲𝗶𝗿 𝗺𝗼𝘁𝗵𝗲𝗿’𝘀 𝘃𝗼𝗶𝗰𝗲… 🤍 Ann’s daughter was just one year old when a severe brainstem stroke took her mother’s ability to speak. For 18 years, Ann knew exactly what she wanted to say. Her body just couldn’t turn those thoughts into words. Then scientists connected her brain to a computer. AI learned to decode the neural signals produced when she attempted to speak – translating them into text, speech and even the facial expressions of a digital avatar. 253 electrodes. Up to 78 words per minute. Her previous communication device: around 14. And perhaps the most human detail: “My daughter was 1 when I had my injury, it’s like she doesn’t know Ann … She has no idea what Ann sounds like.” So researchers used a recording of Ann speaking at her wedding to recreate a synthetic version of her own voice. When Ann heard it, she said: “My brain feels funny when it hears my synthesized voice. It’s like hearing an old friend.” And now, she looks forward to the day when her daughter — who has only known her mother’s computerized voice with a British accent — can hear it too. This isn’t mind reading. It’s neuroscience, AI and a brain-computer interface working together to give someone back something deeply human: a voice. 𝗧𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝘆 𝗱𝗶𝗱𝗻’𝘁 𝗴𝗶𝘃𝗲 𝗵𝗲𝗿 𝗯𝗼𝗱𝘆 𝗯𝗮𝗰𝗸. 𝗜𝘁 𝗴𝗮𝘃𝗲 𝗵𝗲𝗿 𝗮 𝘄𝗮𝘆 𝘁𝗼 𝗯𝗲 𝗵𝗲𝗮𝗿𝗱 𝗮𝗴𝗮𝗶𝗻. 🤍 What technology gives you the most hope for the future? For more TECHNOLOGY WITH HUMAN IMPACT, follow: Lara Sophie Bothur 𝘚𝘰𝘶𝘳𝘤𝘦: 𝘔𝘦𝘵𝘻𝘨𝘦𝘳 𝘦𝘵 𝘢𝘭., 𝘕𝘢𝘵𝘶𝘳𝘦, 2023; 𝘜𝘊𝘚𝘍.

Last week, we explored how robots might move, feel, and understand like humans. Now, we flip the lens and tap into one of the most exciting frontiers in human augmentation: Brain-Computer Interfaces (BCIs). BCIs connect the brain directly to machines, translating neural activity into signals that control computers, devices, or even AI agents. With the rise of Agentic AI, a new possibility is emerging: What if your intentions could become instructions, from brainwaves to prompts, directing AI with intent alone? The most intuitive interface isn’t voice; it’s thought. A Thought-to-Agent Interface (T2A) links your brain activity to an AI Agent in real time, translating mental focus, intention, or emotional state into prompts, actions, or decisions. These are some use-case examples... 🧠 In Work: You're in deep focus. You imagine a slide, your AI Agent starts drafting it. You think of a person; it pulls up your last conversation. 🧠 In Accessibility: For someone unable to speak or type, the interface interprets intent from brain signals and helps control devices, compose messages, or navigate systems. 🧠 In Creativity: A designer imagines a shape, a scene, or a melody, and the AI Agent renders variations in real time, refining the output through guided intent. These are some current research projects... 📚 Meta AI’s Brain-to-Text Decoding: Decodes full sentences from non-invasive brain activity with up to 80% character accuracy, bridging neural intent to digital language. https://lnkd.in/gTEJpa4e 📚 UC Berkeley’s Brain-to-Voice Neuroprosthesis: Translates brain signals into audible speech, restoring naturalistic communication for people with speech loss. https://lnkd.in/g_D3Xeup 📚 Caltech’s Mind-to-Text Interface: Achieves 79% accuracy in translating imagined internal speech into real-time text, enabling seamless brain-to-device communication. https://lnkd.in/gEuVKreq These are some startups to watch... 🚀 Neurable: EEG-based wearables decoding cognitive load & focus in real-time. https://www.neurable.com/ 🚀 OpenBCI: Makers of Galea, a headset combining EEG, EMG, eye tracking, and skin conductance for immersive neural interfacing. https://lnkd.in/girt4PAW 🚀 Cognixion: Brain-powered communication integrated with AR and speech synthesis for non-verbal users. https://www.cognixion.com/ 🚀 Paradromics: High-bandwidth BCI for translating neural activity into speech or system commands for those with severe impairments. https://lnkd.in/giepGKH4 What is a likely time horizon... 1–2 years: Wearable EEG interfaces paired with AI for narrow tasks: adaptive UI, hands-free control, attention-based interaction. 3–5 years: Thought-to-agent pipelines for work, accessibility, and creative tools, personalized to individual brain patterns and cognitive signatures. The future isn’t just AI that understands your prompts. It’s AI that understands you as soon as you think. Next up: Multimodal AI Sensory Fusion (“Glass Whisperer”)

Our latest research is now online, unravelling the dynamic of ultrasound stimulation on the human brain! How does functional connectivity change after #TUS and how is this linked to cognitive changes? By targeting either the inferior frontal cortex (IFC) or the thalamus (THAL), we observed distinct changes in fMRI up to 1h after the stimulation. Testing around 1.5 hours after stimulation, reaction time was 8% faster for the IFC target in a stop-signal-task. Accuracy remained as high as before stimulation. This cognitive change in reduced reaction time was directly linked to biological change (rs-fMRI functional connectivity) and physical change (pressure change at target) as also confirmed by mediation analysis. Altogether, this highlights three points: (1) TUS can lead to long-term (1h+) cognitive benefits (faster reaction time) (2) Cognitive change can be predicted from biological (brain activity) change (3) Biological, and thus cognitive, effect can be inferred from physical change You can find the study here: https://lnkd.in/d-hmVeDV

A high-performance speech neuroprosthesis, developed by Stanford researchers, decodes attempted speech directly from brain activity—restoring a voice to individuals who have lost the ability to speak. Key Findings: 📍Rapid and naturalistic decoding: The system translated neural signals into real-time text at 62 words per minute—nearly 3.5× faster than prior BCI systems. This speed brings decoded communication closer to everyday conversation, offering a major leap in usability and responsiveness. 📍Robust phoneme mapping and vocabulary range: Impressively, the neuroprosthesis operated with a 125,000-word vocabulary—the largest ever used in speech BCI—while maintaining semantic accuracy. Neural representations of phonemes remained intact even years after speech loss, suggesting the brain’s motor-speech pathways are more persistent than previously assumed. 📍Rethinking the neural basis of speech: While traditional models emphasize Broca’s area, this study found that area 6v was more predictive of speech intention. Furthermore, the system successfully decoded both spoken and silently mouthed words, demonstrating that silent articulation retains a reliable neural signature—crucial for fatigue-free, discreet communication. By Willett et al., Nature, 2023 https://rdcu.be/eyFkC Implication: This work marks a major milestone for brain–computer interfaces, bridging neuroscience and assistive technology to restore speech—and reshaping our understanding of the brain’s language architecture. #BrainComputerInterface #Neuroprosthetics #SpeechNeuroprosthesis #Neuroscience #Stanford #ALS #Neurotech #BCI

Generative AI: 𝟭𝟳-𝗬𝗲𝗮𝗿 𝗢𝗹𝗱 𝗕𝗲𝗻 𝗖𝗵𝗼𝗶 𝗖𝗿𝗲𝗮𝘁𝗲𝘀 𝗠𝗶𝗻𝗱-𝗖𝗼𝗻𝘁𝗿𝗼𝗹𝗹𝗲𝗱 𝗣𝗿𝗼𝘀𝘁𝗵𝗲𝘁𝗶𝗰 At just 17, Benjamin Choi revolutionized prosthetic technology with a $300 mind-controlled arm that rivals devices costing $450,000 plus. 𝗧𝗵𝗲 𝗦𝗽𝗮𝗿𝗸 Inspired by a 2012 "60 Minutes" documentary on brain-controlled prosthetics, young Ben was fascinated—but alarmed by the invasive brain surgery and astronomical costs that kept this life-changing technology out of reach 𝗧𝗵𝗲 𝗣𝗮𝗻𝗱𝗲𝗺𝗶𝗰 𝗣𝗶𝘃𝗼𝘁 When COVID shut down his aluminum fuel research lab in 2020, Ben transformed his basement ping-pong table into a makeshift laboratory, working 16-hour days with just a $75 3D printer. 𝗧𝗵𝗲 𝗦𝗰𝗶𝗲𝗻𝗰𝗲 Ben's breakthrough came through developing a revolutionary AI algorithm that interprets brain signals from NON-INVASIVE sensors ▶  External electrodes on forehead and earlobe (no surgery!) ▶  23,000+ lines of custom code across 7 novel sub-algorithms ▶  978 pages of advanced calculus and machine learning ▶  95% accuracy vs. industry standard of 73.8% The story is also marked by: 𝗛𝘂𝗺𝗮𝗻-𝗖𝗲𝗻𝘁𝗲𝗿𝗲𝗱 𝗥𝗲𝘀𝗲𝗮𝗿𝗰𝗵 Ben conducted IRB-approved studies with 6 volunteers, collecting thousands of brainwave data points while participants focused on hand movements—training his AI to decode neural intentions 𝗘𝗻𝗴𝗶𝗻𝗲𝗲𝗿𝗶𝗻𝗴 𝗘𝘅𝗰𝗲𝗹𝗹𝗲𝗻𝗰𝗲 After 75+ design iterations, Ben's prosthetic uses engineering-grade materials that withstand 4+ tons of force, with real-time Bluetooth communication between brain and limb 𝗥𝗲𝗰𝗼𝗴𝗻𝗶𝘁𝗶𝗼𝗻 & 𝗜𝗺𝗽𝗮𝗰𝘁 It's hard to take in the remarkable number of awards this scientist and engineer has achieved, regardless of his age: ▶  Davidson Fellows Laureate ($50K scholarship) ▶  Regeneron Science Talent Search Top 40 Finalist ▶  MIT THINK Scholar ▶  Now researching at Harvard + Johns Hopkins Applied Physics Lab ▶  Published in Journal of Neural Engineering The Real Innovation: Ben proved that cutting-edge neural interfaces don't require invasive procedures or massive budgets—opening doors for millions of amputees worldwide. From third-grader watching documentaries to teenage inventor changing lives. Sometimes the most profound breakthroughs come from asking: What if there's a better way?" Have a great week! #TScottClendaniel #AI #ArtificialIntelligence #Education #GenAI #MachineLearning #Technology #Training

Two people with paralysis, one with ALS, one with a spinal cord injury, typed at 22 words per minute using a brain-computer interface. Not in a research hospital. In their homes. Published this week in Nature Neuroscience by the BrainGate Team. That's the clinical story. Here is the signal. While Neuralink talks high-volume production and human-machine symbiosis, a university team quietly demonstrated what actually matters: a paralyzed person communicating at near-normal speed, reliably, where they live. The gap between spectacle and utility just closed, on the clinical side, not the commercial one. China noticed. Beijing's 15th Five-Year Plan elevates BCI to a strategic "industry of the future" alongside quantum and 6G, targeting world-class firms by 2030. This month, China's NMPA granted Neuracle's NEO implant commercial clearance, while Neuralink's device remains in US clinical trials. More than ten invasive human trials are underway. Pilot provinces already cover BCI treatments under national medical insurance. The US leads on evidence. China is building state-backed infrastructure to commercialise it. Neither has answered the harder question: who pays for neural interfaces when they outperform every assistive device on the market? Disability economics were designed for eye-trackers, not cortical keyboards. The BCI race isn't between Neuralink and its competitors. It's between two governance models, and neither is ready. Sources in comments.