A new technique based on artificial intelligence is capable of translating brain MRI images into specific words and phrases, according to a team of researchers from the university of texas. Although it is still in the preliminary phase, in the future it could help people with paralysis to recover their ability to communicate.
The interface, developed by computational neuroscientist alexander huth, tracks changes in brain blood flow to measure neural activity. The experimental phase of the research consisted of scanning the brains of three participants, each of whom spent listening to 16 hours of podcasts. With the information obtained, a set of maps was created that allowed identify words, phrases and prayers linked to those stories. Scientists also appealed to the platform ChatGPT to anticipate which word would follow the others in the internal discourse of those involved.
Samsung bans its staff from using AI after ChatGPT data leak
In a second part of the experiment, they were shown a movie without sound. At one point, while one of them was watching an animation where a dragon kicked someone to the ground, the system identified what was happening without the need for the viewer to speak.
To improve this technique more data training will be needed and the use of more practical equipmentsuch as portable electroencephalograms. Privacy is an additional difficulty. If the system managed to decode spontaneous thoughts in real time, it would directly be reading minds.
“We need all participants to ensure that this happens in an ethical way,” he asked. Nita Farahanyexpert in bioethics by Duke University. “It can be truly transformative for those who need to get back in touch, but the implications are profound for the rest of us.”
BL J L
You may also like