3 min read
3 comments
Man with paralysis achieves historic feat that could revolutionize the use of artificial intelligence in medicine. Study participant moves robotic arm with just his thoughts.
Researchers at the University of California, San Francisco (UCSF) have created a brain-computer interface that allows a paralyzed man to control a robotic arm using only his thoughts, changing the course of artificial intelligence in medicine. The thought-controlled robotic arm, which operated for a record 7 months without needing adjustments, represents a significant advance in a technology that previously required daily or bi-day recalibration.
Artificial intelligence in medicine
The participant to use the thought-controlled robotic arm, who was paralyzed after a stroke years ago and cannot speak or move, had tiny sensors implanted on the surface of his brain.
Os sensors with Artificial Intelligence capture brain activity when he imagines performing specific movements. Through this system, the man with paralysis was able to grab, release and move objects simply by imagining performing these actions.
- US warns of possibility of Chinese army with genetically modified soldiers and artificial intelligence support, according to official report
- New Chinese company arrives in Brazil and will generate 500 jobs: Realme opens the first factory in Latin America with the capacity to produce up to 20.000 cell phones per day
- Being nice and saying 'please' and 'thank you' to ChatGPT is costing OpenAI thousands of dollars
- The James Webb Space Telescope has identified a spiral galaxy similar to the Milky Way
The secret to success was the discovery of how brain activity changes daily as a person repeats a movement in imagination and learns to perform it in a more refined way.
According to Dr.Karunesh Ganguly, a neurologist and professor at UCSF, this combination of human-to-human learning and artificial intelligence in medicine uses an AI model programmed to adjust to the tiny changes that happen in the brain as the user perfects their imagined movements.
The team of researchers found that while the brain patterns maintained their general shape, their locations shifted slightly each day. This variation explains why other similar devices quickly lost their ability to recognize brain commands.
Understand how the test for a thought-controlled robotic arm works
To train the paralyzed man, scientists had him practice with a virtual robotic arm that provided feedback on the accuracy of his visualizations. Once he mastered the virtual control, he transferred those skills to the real world within a few practice sessions.
The participant was able to move the thought-controlled robotic arm to pick up blocks, rotate them, and move them to new locations.
It was even possible to open a cabinet, remove a glass and place it under a water dispenser, showing a major advance in Artificial Intelligence in medicine. Months later, the paralyzed man was still able to move the thought-controlled robotic arm after a quick 15-minute recalibration to adjust to changes in brain representations that had occurred since he first started using the device.
The university team continues to refine AI models in medicine to make the movements of the thought-controlled robotic arm faster and smoother, as well as planning tests in the home environment.
For people with paralysis, the ability to feed themselves or drink water would represent a significant change in their quality of life. The study, funded by the National Institutes of Health (NIH) and the UCSF Weill Institute for Neuroscience, was published in the journal Cell on March 6, 2025.
About UCSF
The University of California, San Francisco (UCSF) is exclusively focused on the health sciences and aims to promote health worldwide through advanced biomedical research, graduate-level education in the life sciences and health professions, and excellence in patient care.
UCSF Healthcare, which serves as UCSF's flagship academic medical center, includes top-tier specialty hospitals and other clinical programs, and has affiliations throughout the Bay Area.
Miguel Nicolelis was forgotten in the article, do you know why? He is Brazilian and a pioneer in the brain interface binomial. How provincial we are, even when we are not.
There is a lack of support here in Brazil in the area of scientific Biomedicine. Few people know, but Biomedicine is the one truly responsible for the studies and creation of these advanced technologies.
All that was missing was to give due credit to Professor Nicolelis' pioneering work: https://youtube.com/shorts/le4eFufKsQg?si=A2-iuw2-LsYG4U6s