summary: The researchers created an expanded computer interface in the brain, enhancing artificial intelligence, allowing users to control an arm or robotic index with more accurate and speed. The system translates brain signals from EEG records into movement orders, while the AI camera explains the user’s intention in actual time.
In the tests, the participants – including a paralyzed individual – completed the tasks significantly with the help of artificial intelligence, even with the completion of unnecessary procedures without them. The researchers say this penetration can pave the way for safer auxiliary techniques access to people with paralysis or motor disabilities.
Main facts:
- Increased penetration: It combines EEG -based decoding with an artificial intelligence of common autonomy.
- Complete the task faster: Poor participants can complete the tasks that were impossible without the assistance of Amnesty International.
- An alternative to reach: It provides safer and less dangerous solutions compared to gas surgical agriculture.
source: Ucla
UCLa engineers have developed a computer interface system in the brain that can be worn as expanded that use artificial intelligence as a participant pilot to help infer the user’s intention and complete the tasks by moving a roboty or a computer index.
Posted in The intelligence of the nature machineThe study shows that the interface shows a new level of performance in the unintended brain computing interface, or BCI, systems.
This may lead to a set of techniques to help people with limited physical abilities, such as those that suffer from paralysis or nervous conditions, while dealing with things and moving them more easily.
The team has developed a dedicated algorithm for disassembling of the Electrical layout, or EEG – a way to record the electrical activity of the brain – and extract signals that reflect the movement’s intentions.
They paired the signals that were decoded with the camera -based artificial intelligence platform that explain the user’s direction and intentionally. The system allows individuals to complete the tasks much faster than the help of Amnesty International.
“By using artificial intelligence to supplement the computer interface systems in the brain, we aim at least dangerous ways and conquest,” said study commander Jonathan Kaw, Associate Professor of Electrical and Computer Engineering at the University of California College in Samuel.
“Ultimately, we want to develop AI-BCI systems that offer common independence, allowing people with movement disorders, such as paralysis or Als, to restore some independence for daily tasks.”
Surgically cultivated BCI devices and surgically transplanted brain signals into orders, but the benefits currently provide them with risks and costs associated with neurosurgery to plant them.
After more than two decades have passed for the first time, these devices are still limited to small experimental clinical trials. Meanwhile, BCIS showed that it can be worn and other BCIS fewer performance in discovering brain signals.
To treat these restrictions, the researchers tested the new BCI with the help of AI with four-three participants without disabilities in cars and the fourth that paralyzed from the waist to the bottom.
Participants wore an EEG registration headscarf, and researchers used decoding algorithms dedicated to translating these brain signals into computer indicator movements and automatic arm. At the same time, note the artificial intelligence system with a built -in camera that has been decoded and helped the participants complete two tasks.
In the first task, they were directed to move the index on the computer screen to hit eight goals, with the indicator in place for at least half a second. In the second challenge, the participants were asked to activate an automatic arm to move four blocks on a table from their original sites to the specific situations.
All participants have completed both tasks much faster with the help of artificial intelligence. It is worth noting that the paralyzed participant completed the mission of the mechanical arm in about six and a half minutes with the help of artificial intelligence, while without it, he was unable to complete the task.
BCI decoded the electrical brain signals that cod the intended procedures for the participants. Using a computer vision system, artificial intelligence dedicated to user’s New Note-not their eye movements-to direct the indicator and lumps.
“The following steps for AI-BCI systems can include the development of more common spectrums that move automatic weapons at more speed and accuracy, and provide a skilled touch that adapts to the object that the user wants to understand,” said the co-author of the PhD in UCL Electrical Engineering.
“The addition of large -scale training data can also help Amnesty International in cooperating in the most complex tasks, as well as improving the EEG decipher itself.”
The authors of the paper are all members of the Kaw Engineering and Neuroscientist Laboratory, including Sanjun Lee and Epishek Merena, Shu Yan, Brandon McMahman, Brent Gayvioard, Charles Kobashiagawa, Mike Koy, Chang Shi. A member of the Institute of Brain Research at the University of California, Los Angeles, KAO also holds the dates of faculty members in the Department of Computer Science and PhD among departments. A program in neuroscience.
Finance: The research was funded by the National Institutes of Health and the Center for Science for Humanity and Artificial Intelligence, a cooperation between the University of California in Los Angeles and Amazon. The UCla Technology Development Group has applied for a patent for AI-BCI technology.
On these news, research in the field of nervous technology and AI
author: Christine Wei Lee
source: Ucla
communication: Christine Wei-Lee-California University
image: The image is attributed to news of neuroscience
The original search: Closed access.
“The brain’s interface control and selection with artificial intelligence coilotsBy Jonathan Cao, and others. The intelligence of the nature machine
a summary
The brain’s interface control and selection with artificial intelligence coilots
Dismantling of the brain and fractures (BCIS) nervous signals to help people with paralysis to move and communicate.
Even with important progress in the past two decades, BCIS is a major obstacle to clinical feasibility: BCI should be highly expensive and risks.
To significantly increase the performance of BCI, we use joint self -rule, as Copilots collaborates artificial intelligence (AI) with BCI users to achieve task goals.
We explain this AI-BCI in non-invasive BCI electrical layout signals.
First, we contribute to the mixed adjustable decoding approach by using a fodder nervous network and the renewal kalman candidate, allowing healthy users and participating with paralysis to control computer indicators and automatic weapons through separate electrical photography signals.
Then we design two AI Copilots to help BCI users in the index control task and the task of selecting the automatic arm.
AI-BCIS, who enables the participant with paralysis to achieve the performance of 3.9-Times, explain the target hit rate while controlling the index and controlling an automatic arm to move in successive random blocks into random sites, a task they could not do without AI Copilot.
With the improvement of Copilots AI, the designed BCIS may achieve a higher performance independence.
2025-09-01 17:30:00
