I am a postdoctoral researcher in the Department of Computing and Software at McMaster University. My current work is part of an NIH-funded project to develop an accessible, reproducible analysis pipeline for neurocurrent response functions (NCRFs). Like temporal response functions (TRFs), NCRFs estimate the linear neural response to continuous stimuli, but they do so directly in source space, jointly performing source localization and response estimation from magnetoencephalography (MEG) data. Using existing MEG datasets and computational models, my research examines how the human brain transforms continuous speech into linguistic meaning.
I completed my PhD in Cognitive Robotics at the Okinawa Institute of Science and Technology Graduate University (OIST) in 2024, an MS in Applied Cognition and Neuroscience at the University of Texas at Dallas in 2014, and a BTech in Biotechnology at Anna University in 2012. My PhD work focused on brain-inspired neural networks that let robots learn language and action together. I developed a network architecture based on the Free Energy Principle, trained it with multimodal data from an arm robot, and showed how it combines familiar concepts in new ways, a hallmark of compositionality. Earlier in my career, I studied ghrelin and related questions in metabolic neuroscience, including exercise endurance, feeding, and glucose regulation.
My research asks how biological and artificial systems come to understand language. I analyze existing MEG datasets in which participants listen to natural, continuous speech, standardizing them in the Brain Imaging Data Structure (BIDS) format and modeling them with Eelbrain and NCRFs. These models estimate how acoustic and linguistic features of the speech drive neural activity across cortical regions over time. My broader goal is to characterize how the brain maps external signals onto abstract linguistic representations and composes them into larger structures.
Search for Prasanna Vijayaraghavan's papers on the Research page