Researchers Develop AI That Reconstructs Images From Brain Activity

Researchers have successfully developed an AI system that can reconstruct images viewed by humans by analysing brain activity recorded through functional magnetic resonance imaging (fMRI) scans. The technology works by processing neural signals captured during the scanning process and converting them back into visual representations of what the subject was observing.

The approach represents a significant advance in understanding how the human brain processes visual information. By mapping the relationship between patterns of neural activation and the visual stimuli that triggered them, the AI can learn to predict and recreate images based solely on brain activity data. This opens new possibilities for studying visual perception at a fundamental level.

The development builds on decades of neuroscience research into how visual cortex regions encode information about the world around us. The AI system leverages machine learning to identify the complex associations between specific patterns of brain activation and corresponding visual features, allowing it to generate images that closely match what a person's brain was processing during scanning.

While the technology remains in research stages, scientists suggest it could eventually have practical applications in neuroscience, helping researchers better understand visual perception disorders and neurological conditions affecting sight. The work also raises important questions about privacy and the interpretation of neural data as the technology becomes more refined.