My early work in the UK which formed the basis of my doctorate was on the neural organization of the pupillary light reflex pathway. Following that I worked with Gian Poggio on the response properties of neurons in the primate foveal striate cortex in the Department of Physiology at Johns Hopkins University Medical School Baltimore. On my return to the UK I set up a laboratory that pioneered the application of pharmacological tools and particularly micro-iontophoretic application of drugs to the dissection of the functional organization of the central visual system. This led to new insights into the role of GABA mediated inhibition in the thalamus and cortex and the role that excitatory amino acid receptors play in the transfer of information in the lateral geniculate nucleus. Our primary research focus then switched to the role of feedback pathways and the way this is linked to the influence of stimulus context. Building on this, and our earlier techniques, the current work from my group is concerned with the synaptic mechanisms of vision and in particular with the interplay between feed forward and feedback systems in visual processing. Recent publications (J Neurophysiol 2001 and 2002) show how the context of a stimulus can transform the way neurons in the primary visual cortex respond to local features and argue that feedback interactions constantly update local processing to rephrase it in terms of the integration and hypotheses abstracted at the higher levels. A paper in Nature Neuroscience (2006) using pharmacological manipulation of feedback cells provides the first demonstration of the functional logic of the alignment of feedback connections from layer 6 of the visual cortex in the thalamus, showing that the connections are orientation aligned and phase reversed. This is a mirror image of the Hubel and Wiesel model for feed forward connections. Another paper in PNAS (2007) shows that the reciprocal interaction between visual cortex and thalamus substantially enhances the precision in the way the thalamic neurons signal the presence of moving contours. In this sense the pattern of the thalamic neurons response to moving visual stimuli is an emergent property of the interaction between thalamus and cortex. The ongoing program over the next five years is concerned with the role of feedback in motion processing and the way this categorizes and integrates the different components of the visual world in motion enabling both fine scale segmentation and large scale integration. We find a clear track of influence from the cortical motion area MT/V5 on the response properties of primate thalamic cells to all classes of visual stimuli. All this indicates that the Bayesian priors embedded in the higher visual cortical areas, such as MT/V5 change, via cascaded feedback, the behaviour of the earliest thalamic mechanisms. It argues that we should view the system as a circuit in which the influence of all levels are integrated in a moment by moment fashion with little or no real delay.

We also have and are making a significant contribution to the translation of this knowledge into clinical research and benefit to the patient (see example publications in PNAS, Guo et al., 2007, Cordeiro et al., 2004) and have a current project with Robert MacLaren using a primate model to develop and evaluate improved strategies for treating age related macular disease.