Content on this page requires a newer version of Adobe Flash Player.

Get Adobe Flash player

About Me - Dr Peter Garrard

I studied medicine as a second degree (previously an Oxford classicist), and qualified from Bristol University in 1990. After house jobs in Edinburgh (surgery) and Yeovil (medicine), I completed general medical training in London, and then specialist training in neurology 2000.  During neurology training I was a Medical Research Council clinical training fellow in Cambridge, where I completed a PhD on language abnormalities in Alzheimer's and other dementias.

My first consultant job was at the National Hospital for Neurology and Neurosurgery, which I held while following up my doctoral research as an MRC Clinician Scientist Fellow at the Institute of Cognitive Neuroscience.  After six years at University College London I was offered a senior position at the University of Southampton School of Medicine.  I took up my current post as Reader in Neurology St. George's in March 2010.

My research focuses on:

  • Early onset Alzheimer's disease and other dementias
  • Language disorders in dementia
  • Language abnormalities in texts written by dementia sufferers
  • Brain imaging of dementia and its relationship to language difficulties

Techniques used in my research include:

  • Experimental neuropsychology (measuring the cognitive abilities of brain injured people, including those with dementia)
  • Magnetic resonance (MR) brain imaging
  • Computerised approaches to text analysis, such as digital stylometry and latent semantic analysis
  • Event related potentials:  patterns of cortical activity timelocked with cognitive tasks, detected during processing by scalp electrodes.  Readings provide rough spatial localisation, but excellent temporal resolution (in comparison with fMRI).  Precise spatial and temporal resolution can occasionally be achieved from patients with intracranial electrodes implanted prior to epilepsy surgery.
  • Connectionist modeling: aspects of learning and cognition simulated using distributed representations and learning algorithms implemented in a network of neuron-like units with modifiable interconnections.
  • Connectionist modeling  (simulations of aspects of human cognition)