A new study suggests AI-assisted brain imaging is moving beyond science fiction. Norwegian neuroscientist Kai J. Miller at the University of Oslo calls the technology remarkable. His team used artificial intelligence to decode visual information directly from brain scans.
The research builds on earlier work in neural decoding. Previous methods required subjects to focus on specific images for extended periods. The new approach analyzes spontaneous brain activity without such constraints. Scientists recorded brain signals using functional MRI and trained algorithms to reconstruct seen images.
Miller emphasizes the technology’s potential for medical applications. It could help patients with locked-in syndrome communicate by interpreting their thoughts. The study also examined how different brain regions process visual information. Results showed high accuracy in reconstructing simple shapes and colors.
Ethical concerns emerge alongside technical progress. Researchers acknowledge risks of misuse in surveillance or privacy violations. Norway’s data protection authority has been consulted during the project. The team stresses the need for strict ethical guidelines before wider deployment.
While still in early stages, the findings mark a significant step. Other labs have attempted similar experiments with mixed results. This study stands out for its higher success rate and simpler requirements. The work appears in the journal Nature Neuroscience.
Source: digi.no