Dr Louise Whiteley
Gatsby Computational Neuroscience Unit, UCL
17 Queen Square
Tel: +44 207 679 1172
Email: lewhiteley (at) gmail.com
My recently completed PhD thesis, which was supervised by Dr Maneesh Sahani can be downloaded here.
I am currently deputy editor of The Liberal magazine, and am studying for an MSc in Science Communication at Imperial College London. Along with Mr James Wilkes and Prof. Geraint Rees, I'm also working on a new radio play and event series exploring how images of the brain affect how we think about the mind, supported by a Wellcome Trust public engagement grant and LCACE funding. For more details, see my homepage.
Bayesian optimality and probabilistic population codes
I am interested in the hypothesis that the brain represents external and internal sources of uncertainty or "noise", and that it computes with these uncertainties according to Bayesian probability theory. My PhD project used human psychophysical experiments to provide evidence for this hypothesis, and I also conducted preliminary investigations into what kind of neural population codes might underlie the observed behaviour. In particular, I am interested in exploring the ubiquity of such processing, and how the simple codes so far proposed can be extended to more complex cases.
Anatomical correlates of optimal decision making
There is little understanding of the anatomical basis of an optimal decision, which involves the integration of sensory uncertainty with external and internal loss functions, prior beliefs, and subjective preferences for risk or ambiguity. I have contributed to an fMRI study investigating where in the stream of processing from visual analysis to motor act value might affect processing, in collaboration with Mr Stephen Fleming, Dr Oliver Hulme, and Prof. Ray Dolan at the Wellcome Trust Centre for Neuroimaging.
A probabilistic framework for top-down attention
As part of my PhD project, I also worked on a Bayesian computational framework for top-down attention. Dynamical neural models have successfully described 'bottom-up' attentional selection and some features of the behaviour and physiology associated with 'top-down' attention. However, an explanation of why attention is needed, as well as precisely what computational role it plays has proved elusive. One possible reason for this is that the various phenomena grouped under the rubric of attention may, in fact, be heterogenic. Indeed, although the concept of a limited processing resource is often invoked to explain and unify these phenomena, attempts to give this concept a concrete definition in cognitive or anatomical terms have failed. We aim to provide a normative computational account of the nature of the limited resource, why it is limited, and how attention makes tractable the computationally innaccessible operations that result. In brief, the limitation is in representing belief distributions over multiple spatial or featural quantities, meaning that correlations in the resulting joint distributions are often neglected. In this framework, attention provides probabilistic "hypotheses" which dynamically evolve in order to locally refine these impoverished approximations to the true joint distribution. A paper and related book chapter are currently in preparation.
Modelling saliency responses in fMRI
I am currently collaborating on a fMRI project investigating the role of the posterior thalamus in the salience computations thought to underlie bottom-up allocation of attention. We are using high resolution fMRI to image the brains of human volunteers whilst they reported the salience of competing stimuli, and will use the data to explore the spatial response profiles of voxels in the pulvinar.
Current External Collaborators:
Dr Oliver Hulme and Dr Stuart Shipp at the Institute of Opthalmology, UCL.
Mr Steve Fleming, Prof. Ray Dolan, Prof. Karl Friston, and Prof. Chris Frith at the Wellcome Trust Centre for Neuroimaging, UCL.
Whiteley, L., & Sahani, M. (in preparation). A Bayesian framework for selective attention.
Whiteley, L., & Sahani, M. (2008). Implicit knowledge of visual uncertainty is used to guide statistically optimal decisions. Journal of Vision, 8, 1-15.
Whiteley, L., Spence, C., & Haggard, P. (2008). Visual processing and the bodily self. Acta Psychologica, 127, 129-136.
Hulme, O.J., & Whiteley, L. (2007). The "mesh" as evidence - model comparison and alternative interpretations of feedback. Behavioural and Brain Sciences, 30, 505-506.
Yarrow, K., Whiteley, L., Rothwell, J.C.E., & Haggard, P. (2007). Biases in the perceived timing of perisaccadic perceptual and motor events. Perception & Psychophysics, 68, 1217-1226.
Yarrow, K., Whiteley, L., Rothwell, J.C.E., & Haggard, P. (2006). Spatial consequences of bridging the saccadic gap. Vision Research, 46, 545-555.
Whiteley, L., Kennett, S., Taylor-Clarke, M., & Haggard, P. (2004). Facilitated processing of visual stimuli associated with the body. Perception, 33, 307-314.
last updated 29-01-2009