TITLE: Examining the processing properties of the Cerebral Cortex
SPEAKER: Fawad Jamshed & Andrew Thwaites, Neurolex Group (Psychology Department, Cambridge University)

ABSTRACT:

The brain is a notoriously complex mechanism, and understanding how it works is of importance to a large number of fields, including medicine and artificial intelligence. This work is hampered by two major difficulties. The first difficultly is that the brain is capable of very diverse behaviours. This means that brain-related hypotheses have to be suitably complex if they are going to explain brain behaviour adequately. As a result, the time spent constructing even a single hypothesis can sometimes take years.

The second difficulty is that the majority of brain-related hypotheses, whatever their form, are difficult to test. This is because of the immense difficulty of measuring anything in the in vivo brain. The cerebral cortex (the part of the brain thought to be responsible for cognition) is constructed of billions of neurons encased in a skull, and measurements of dynamic neuronal activity are short-lived and difficult to extract. For those few dynamic features that can be calculated or quantified (i.e. using EMEG or ECoG etc.), the subsequent measurements are invariably bound by assumptions, caveats and complications (including high signal-to-noise ratios and spatial correspondence problems).

Neurolex looks at hypotheses relating to the functional properties of the brain’s language processing system. In this talk we will describe how Neurolex attempts to circumnavigate the problems described above, using modern neuroimaging and pattern analysis techniques, including Representational Similarity Analysis (RSA) and neurokymography (NKG).

http://www.neurolex.psychol.cam.ac.uk/research/methods