Output details

EEG decoding of semantic category reveals distributed representations for single concepts

<28> This was the first work to decode the semantics of single concepts from EEG brain recordings, published in the premier journal for neuroscience of language, and cited by researchers at Max Planck and Stanford. Previous work had shown effects only over groups of concepts, making results vulnerable to perceptual confounds. In a collaboration between Trento and Essex, we adapted machine learning methods from brain-computer-interfaces (BCI) and applied them in the neuroscience of higher cognition. The analysis of single stimulus presentations enabled us to exclude arbitrary confounds for the first time, conclusively detecting object understanding rather than image perception.