Models of Cortical Processing


One of the most worthwhile tasks in any area of science is the persuit of a general model. The history of physics is the history of such endeavours. In biology, the theory of evolution and the genetic code have proved to be formidable examples that match anything in the natural sciences. As neuroscientists, we have to admit to a far less impressive record: the neuron doctrine and the Hodgkin-Huxley model of the action potential, perhaps? The cerebral cortex remains without its general theory, but there are many who believe that a general theory is possible. Our own research has attempted to unify many different strands of investigation at the level of cortical microcircuits. Our 'canonical microcircuit for neocortex' first published in 1989, has been a very important focus for our continuing experimental and theortical work, and that of others. With such a clear and simply stated hypothesis, we have been able to colour in the outline sketch that was the original circuit. These experimental and theoretical studies have added great weight to the hypothesis. The strength of the detailed anatomical work from our group in particular, has been most persuasive and has compelled others to take our hypothesis seriously in the design and interpretation of their own experiments. Hence 1996 saw the publication of articles, experimental and theoretical, and news and views items and commentaries that centred on the core issues of our hypothesis: that cortical circuits are organised in recurrent excitatory and inhibitory pathways and that that this organisation leads to a number of important emergent properties that seem to be hallmarks of cortical processing. The system is robust in the face of noise, the circuit performs a version of gain control, it is inherently dynamic, and explains in an analytic way a number of puzzling disparities in the literature of visual physiology.

While the interference and delays in a research program caused by a move is always underestimated, the speed at which the experimental program was restarted was a tribute to all those scientists involved. In restarting we have been able to begin new projects, both in the detailed structural work being done on the cortical microcircuits, particularly in the primate, and in the physiological work, which has focussed on the cat in vivo and the rat in vitro. The primate work is presently focussed on area 17, which remains a formidable challenge. However, it is also the ideal candidate for studying important aspects of development and plasticity, which is being carried out as a collaborative HFSP project at the light and ultrastructural level. However, if the 'canonical microcircuit' hypothesis is to be tested then we have to make comparative studies of other key cortical areas. This has begun in the primate with collaborative work on the area MT. This area has been much studied physiologically in behaving monkeys, but nothing is known of its underlying microcircuits. Also, in the rodent, we have begun physiological work on the somatosensory cortex and will extend this to thoeretical and anatomical work. The physiological work on the cat has begun an entirely new project on the complex topic of contrast adaptation and gain control. This work has been very successful and has opened whole new areas for exploration, both experimentally and theoretically. The link between the theoretical work and the experimental physiology is developing well, but much more needs to be done if we are to achieve the goal of a unifying theory. It is already evident that, as hoped, important links are developing between scientists from different disciplines, both within the Institute and outside. As internal knowledge grows, so will the confidence to ask larger questions and perform even more searching experimental and theoretical work.