Jack Gallant
|
Title of the Talk: Decoding the human brain
Abstract: The brain contains signals related to all aspects of sensory processing, motor responses, language, and complex cognitive and decision processes. A device that could successfully decode brain signals would therefore have many useful brain-computer interface applications. However, current approaches to brain decoding are weak; they recover only a miniscule fraction of the information available. We have developed a new approach brain decoding that markedly improves the quality of brain decoding. Our approach does not focus on the decoding problem directly. Instead, we construct statistical models that describe how perceptual and cognitive signals are encoded in brain activity measurements. (Because these encoding models reveal how the brain represents sensory, motor and cognitive signals, they are crucially valuable for basic neuroscience.) |
 |
We then use a Bayesian framework to invert these encoding models in order to decode brain activity. Our results show that this approach accurately classify, identify and reconstruct the structural and semantic aspects of perceptual experiences, and subjective mental states such as visual imagery. Because the framework is quite general, our approach can facilitate decoding of many different sensory, motor and cognitive states. The future of brain decoding technology will depend on the availability of appropriate computational models, and on progress in developing new methods for recording brain signals.
About the Speaker: Jack Gallant is Professor of Psychology at the University of California at Berkeley, and is affiliated with the graduate programs in Bioengineering, Biophysics, Neuroscience and Vision Science. He received his Ph.D. from Yale University and did post-doctoral work at the California Institute of Technology and Washington University Medical School. His research program focuses on constructing quantitative computational models that accurately describe how the brain encodes information during natural tasks, and to use these models to decode information in the brain in order to reconstruct mental experiences. This computational framework can be used to understand and decode brain activity measured by different methods (e.g., functional MRI, NIRS, EEG or ECOG), and in different modalities (i.e., vision, audition, imagery and so on).
