A brain scanner can now reveal exactly what information is being processed—not just which brain regions light up, but the specific visual details a person is using to recognise a face, judge an emotion, or decide a gender. Current brain imaging methods show *where* activity happens, but not *what information* the brain is actually computing. This leaves a critical gap: cognitive neuroscience lacks a rigorous way to reverse-engineer the dynamic flow of fine-grained information from sensory input to decision. The researcher has developed a method that, like Hubel and Wiesel’s classic work on visual cells, dissects complex visual stimuli into tiny information components and tracks how brain networks code and transfer each one across time. If successful, this project will provide the first detailed map of how the human brain processes visual information during real-world tasks—face detection, identity recognition, social trait judgments—across the lifespan. The work is fundamental science: it aims to build the interpretive framework for a new field of “brain algorithmics.” Deeper understanding of how healthy brains compute could eventually inform diagnostics for conditions where information processing goes awry, such as prosopagnosia or dementia, but the immediate payoff is a rigorous, mechanistic account of cognition itself.
View original technical description
Information processing is a pervasive assumption of the most influential models in cognitive neuroscience. For example, in predictive coding, predictions imply explicit knowledge of the detailed information propagated down the visual hierarchy. Likewise, categorical decisions imply the successful match of sequentially accrued, sensorily coded information (the evidence) with memorized information serving as decision criteria (categorical knowledge). Consequently, one of the most pressing developments in cognitive neuroimaging is the development of a brain algorithmics, to reverse engineer the actual information that brain networks dynamically process between stimulus onset and behaviour. The cornerstone of brain algorithmics is the existence of rigorous methods to reverse engineer from brain data the dynamic processing of fine-grained information. Groundbreaking progress in visual neuroscience followed Hubel and Wiesels seminal research on the receptive fields of simple and complex cells. Key to this success was their insight to dissect thevisual input into fine-grained information components and measure how brain cells responded to each. From themapping between information components and cell responses, they inferred the information processing function of the cells. Extending this approach to visual cognition is challenging because stimuli aretypically highdimensional. I have developed methods that dissect these stimuli into fine-grained components. Much like Hubel and Wiesel, I can now quantify where, when and how MEG network nodes code and transfer these information components. Consequently, I aim to apply this unique interpretive framework to understand the information processing functions underpinning visual categorisations. Specifically, I will study how different tasks (i.e. face detection, identity recognition, categorisation of gender, age, emotion, social traits) applied to the same face modulate the dynamic processing of information in brain networks across the lifespan. My approach will provide an unmatched level of interpretation of the coding of brain signals, the function of network nodes and of the information flow.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know