Proceedings of The Physiological Society

Cardiff University (2009) Proc Physiol Soc 17, PC13

Poster Communications

Robotic implementation of inhibition of return; potential insight into the biological equivalent.

S. McBride1, M. Huelse1, M. Lee1

1. Department of Computer Science, Aberystwyth University, Abersytwyth, United Kingdom.


Inhibition of return (IOR) refers to the suppression of stimuli (object and events) processing where those stimuli have previously (and recently) been the focus of spatial attention. In this sense, it forms the basis of attentional (and thus visual) bias towards novel objects. Although the neural mechanism underpinning IOR is not completely understood, it is well established that the dorsal frontoparietal network, including frontal eye field (FEF) and superior parietal cortex are the primary structures mediating its control. These are some of the many modulatory and affecting structures of the deep superior colliculus (optic tectum in non-mammals), the primary motor structure controlling saccade. Although visual information from the retina starts at the superficial superior colliculus, and there are direct connections between the superior and deep layers (Stein and Meredith, 1991) the former cannot elicit saccade directly (Casagrande et al., 1972). This information has to be subsequently processed by a number of cortical and sub-cortical structures that place it 1) in context of attentional bias within egocentric saliency maps (posterior parietal cortex) (Gottlieb, 2007), 2) the aforementioned IOR (Stein et al., 2002), 3) overriding voluntary saccades (frontal eye fields) (Stein and Meredith, 1991) and 4) basal ganglia action selection (McHaffie et al., 2005). Thus, biologically there exists a highly developed, context specific method for facilitating the most appropriate saccade as a form of attention selection. One of the main problems to overcome in constructing an IOR system is the accurate mapping of the retinotopic space to the global egocentric space i.e. foveated objects within a retinotpoic map must be logged within a global egocentric map to allow subsequent comparison with peripheral retinatopic information. The lateral intraparietal (LIP) region is the primary candidate brain region for this process, given its position in modulating the transfer of visual information from superficial to deep superior colliculus. We present here a working robotic model of an IOR system in the context of static objects using camera pan and tilt information (equivalent to head position) to create a global visual memory. The model potentially provides additional biological insight into the types of information transform and transfer that must take place for an accurate IOR system to exist.

Where applicable, experiments conform with Society ethical requirements