Nutriovia

He was able to grasp, move and drop objects simply by imagining himself by performing the actions.

The device, known as the brain interface (BCI), worked for a record of 7 months without having to be adjusted. So far, these devices have only worked for a day or two.

The BCI is based on a model of AI which can adapt to the small changes that occur in the brain as a person repeats a movement – or in this case, an imagined movement – and learns to do it in a more refined way.

“This mixture of learning between humans and AI is the next phase of these cerebral-computer interfaces,” said the neurologist, Karunesh Ganguly, MD, PHD, professor of neurology and member of the UCSF Weill Institute for neuroscience. “This is what we need to achieve a sophisticated and realistic function.”

The study, funded by the National Institutes of Health, appears on March 6 in Cell.

The key was the discovery of the way in which activity moves into the brain day by day as a participant in the study imagining several times by making specific movements. Once the AI ​​has been scheduled to take these quarters of work into account, it worked for months at a time.

Location, location, location

Ganguly studied how models of brain activity in animals represent specific movements and saw that these representations have changed day by day as the animal learned. He suspected that the same thing happened in humans, and that is why their BCIS has so quickly lost the ability to recognize these models.

Ganguly and neurology researcher Nikhilesh Natraj, PhD, worked with a study participant who had been paralyzed by a stroke of years earlier. He could neither speak nor move.

He had tiny sensors located on the surface of his brain which could accelerate brain activity when he imagined in move.

To see if his brain models have changed over time, Ganguly asked the participant to imagine moving different parts of his body, such as his hands, feet or head.

Although he cannot really move, the participant’s brain could still produce the signals of a movement when he imagined himself by doing so. The BCI has recorded the representations of the brain of these movements through the sensors of its brain.

The Ganguly team found that the form of representations in the brain had remained the same, but their locations moved slightly day by day.

From virtual to reality

Ganguly then asked the participant to imagine having simple movements with his fingers, hands or thumbs in two weeks, while the sensors recorded his brain activity to cause AI.

Then the participant tried to control an arm and a robotic hand. But the movements were still not very precise.

Thus, Ganguly had the practice of the participants on a virtual robot arm which gave him comments on the accuracy of his visualizations. Finally, he obtained his virtual arm to do what he wanted him to do.

Once the participant began to train with the real robot arm, he only took a few training sessions so that he transfers his skills in the real world.

He could make the robotic arm pick up blocks, turn them and move to new places. He was even able to open a wardrobe, take out a cup and hold it to a water distributor.

Months later, the participant was able to control the robotic arm after a 15 -minute “update” to adapt to the way in which his movement representations had derived since he had started using the device.

Ganguly now refines AI models so that the robotic arm moves faster and more easily, and plans to test the BCI in a family environment.

For people with paralysis, the ability to eat or have a glass of water would be to change your life.

Ganguly thinks it is at hand.

“I am very convinced that we have learned to build the system now and that we can operate this work,” he said.

He was able to grasp, move and drop objects simply by imagining himself by performing the actions.

The device, known as the brain interface (BCI), worked for a record of 7 months without having to be adjusted. So far, these devices have only worked for a day or two.

The BCI is based on a model of AI which can adapt to the small changes that occur in the brain as a person repeats a movement – or in this case, an imagined movement – and learns to do it in a more refined way.

“This mixture of learning between humans and AI is the next phase of these cerebral-computer interfaces,” said the neurologist, Karunesh Ganguly, MD, PHD, professor of neurology and member of the UCSF Weill Institute for neuroscience. “This is what we need to achieve a sophisticated and realistic function.”

The study, funded by the National Institutes of Health, appears on March 6 in Cell.

The key was the discovery of the way in which activity moves into the brain day by day as a participant in the study imagining several times by making specific movements. Once the AI ​​has been scheduled to take these quarters of work into account, it worked for months at a time.

Location, location, location

Ganguly studied how models of brain activity in animals represent specific movements and saw that these representations have changed day by day as the animal learned. He suspected that the same thing happened in humans, and that is why their BCIS has so quickly lost the ability to recognize these models.

Ganguly and neurology researcher Nikhilesh Natraj, PhD, worked with a study participant who had been paralyzed by a stroke of years earlier. He could neither speak nor move.

He had tiny sensors located on the surface of his brain which could accelerate brain activity when he imagined in move.

To see if his brain models have changed over time, Ganguly asked the participant to imagine moving different parts of his body, such as his hands, feet or head.

Although he cannot really move, the participant’s brain could still produce the signals of a movement when he imagined himself by doing so. The BCI has recorded the representations of the brain of these movements through the sensors of its brain.

The Ganguly team found that the form of representations in the brain had remained the same, but their locations moved slightly day by day.

From virtual to reality

Ganguly then asked the participant to imagine having simple movements with his fingers, hands or thumbs in two weeks, while the sensors recorded his brain activity to cause AI.

Then the participant tried to control an arm and a robotic hand. But the movements were still not very precise.

Thus, Ganguly had the practice of the participants on a virtual robot arm which gave him comments on the accuracy of his visualizations. Finally, he obtained his virtual arm to do what he wanted him to do.

Once the participant began to train with the real robot arm, he only took a few training sessions so that he transfers his skills in the real world.

He could make the robotic arm pick up blocks, turn them and move to new places. He was even able to open a wardrobe, take out a cup and hold it to a water distributor.

Months later, the participant was able to control the robotic arm after a 15 -minute “update” to adapt to the way in which his movement representations had derived since he had started using the device.

Ganguly now refines AI models so that the robotic arm moves faster and more easily, and plans to test the BCI in a family environment.

For people with paralysis, the ability to eat or have a glass of water would be to change your life.

Ganguly thinks it is at hand.

“I am very convinced that we have learned to build the system now and that we can operate this work,” he said.

He was able to grasp, move and drop objects simply by imagining himself by performing the actions.

The device, known as the brain interface (BCI), worked for a record of 7 months without having to be adjusted. So far, these devices have only worked for a day or two.

The BCI is based on a model of AI which can adapt to the small changes that occur in the brain as a person repeats a movement – or in this case, an imagined movement – and learns to do it in a more refined way.

“This mixture of learning between humans and AI is the next phase of these cerebral-computer interfaces,” said the neurologist, Karunesh Ganguly, MD, PHD, professor of neurology and member of the UCSF Weill Institute for neuroscience. “This is what we need to achieve a sophisticated and realistic function.”

The study, funded by the National Institutes of Health, appears on March 6 in Cell.

The key was the discovery of the way in which activity moves into the brain day by day as a participant in the study imagining several times by making specific movements. Once the AI ​​has been scheduled to take these quarters of work into account, it worked for months at a time.

Location, location, location

Ganguly studied how models of brain activity in animals represent specific movements and saw that these representations have changed day by day as the animal learned. He suspected that the same thing happened in humans, and that is why their BCIS has so quickly lost the ability to recognize these models.

Ganguly and neurology researcher Nikhilesh Natraj, PhD, worked with a study participant who had been paralyzed by a stroke of years earlier. He could neither speak nor move.

He had tiny sensors located on the surface of his brain which could accelerate brain activity when he imagined in move.

To see if his brain models have changed over time, Ganguly asked the participant to imagine moving different parts of his body, such as his hands, feet or head.

Although he cannot really move, the participant’s brain could still produce the signals of a movement when he imagined himself by doing so. The BCI has recorded the representations of the brain of these movements through the sensors of its brain.

The Ganguly team found that the form of representations in the brain had remained the same, but their locations moved slightly day by day.

From virtual to reality

Ganguly then asked the participant to imagine having simple movements with his fingers, hands or thumbs in two weeks, while the sensors recorded his brain activity to cause AI.

Then the participant tried to control an arm and a robotic hand. But the movements were still not very precise.

Thus, Ganguly had the practice of the participants on a virtual robot arm which gave him comments on the accuracy of his visualizations. Finally, he obtained his virtual arm to do what he wanted him to do.

Once the participant began to train with the real robot arm, he only took a few training sessions so that he transfers his skills in the real world.

He could make the robotic arm pick up blocks, turn them and move to new places. He was even able to open a wardrobe, take out a cup and hold it to a water distributor.

Months later, the participant was able to control the robotic arm after a 15 -minute “update” to adapt to the way in which his movement representations had derived since he had started using the device.

Ganguly now refines AI models so that the robotic arm moves faster and more easily, and plans to test the BCI in a family environment.

For people with paralysis, the ability to eat or have a glass of water would be to change your life.

Ganguly thinks it is at hand.

“I am very convinced that we have learned to build the system now and that we can operate this work,” he said.