William D. Marslen‐Wilson

The paper identifies a distributed neural substrate for early speech interpretation using MEG data and Representational Similarity Analysis, showing simultaneous processing of phonology, lexical form, and semantics.

This study uses BERT to model incremental speech comprehension and finds that coherent interpretations emerge through the integration of linguistic and non-linguistic constraints in bilateral brain regions.

This study compares speech recognition representations in artificial neural networks and human brains, finding similarities and using brain data to improve machine models.

This study investigates how dopamine levels in Parkinson’s disease patients affect sensory attenuation during voluntary movements, finding a negative correlation with motor severity and a positive correlation with dopamine dose.

This study examines the associations between cardiovascular risk factors and white matter structure in aging, finding that systolic blood pressure, diastolic blood pressure, and heart rate independently impact white matter health.

The paper shows that physical activity preserves frontal white matter integrity and reduces age-related slowing in a population-based sample.

This paper compares dynamic brain states during speech comprehension with dynamic machine states in ASR systems, finding correspondence between neural patterns and phonetic models.