Raymond J Dolan

How did you become interested in medicine? Why neuropsychiatry?

It is always difficult to answer such a question but one thing is that whenever I was sick as a child I felt a certain awe when attended by a doctor. I also suffered with middle ear problems and was exposed to medical specialists who struck me as having extraordinary skill. Why I went into psychiatry subsequently reflects a number of things. Actually, I remember being asked this at the interview for my first job in psychiatry. When I could not give a response, one of the panel noted that I had won a gold medal in ENT as a medical student and thought perhaps the common denominator was an interest in the head! The more plausible reason is that psychiatry seemed the most intellectually interesting part of medicine. Neuropsychiatry, as a discipline within psychiatry, was attractive as it seemed the most grounded in neurobiology, and that was very attractive to me.

Who were the people who inspired you in your career, and what advice of theirs do you remember today?

The people who most inspired me have been the evolutionary biologist Ernst Mayr and, of course, Charles Darwin. In psychiatry, my contacts with Alwyn Lishman early in my career was very important. Also, my colleague here in the Centre, Karl Friston, who incidentally started out as a psychiatrist has had a big influence on me. Karl’s advice, which I believe he received from Michael Gelder who used to be professor of Psychiatry in Oxford, is to always ask the right question. The wisdom goes something like this – “many things are interesting, but only a few are important”.  

Which technique or technology had a profound influence on your career?

Neuroimaging has been the technique that has had the most profound influence on my career. When I entered into psychiatry all we had was a CT scan. The idea that you could read out brain activity in parallel with cognitive processes was just a pipe dream. I was extraordinarily fortunate to be part of a group of people based at the Hammersmith Hospital (MRC Cyclotron Unit) who were in the vanguard of using positron emission tomography (PET) to map cognition, and subsequently being part of the group that set up the Centre which I now direct, the Wellcome Trust Centre for Neuroimaging (WTCN).

You have previously said that the big breakthroughs will come when psychiatry fully embraces neuroscience…what are the early signs of this?

I do think there are early signs of breakthroughs but these are still at the conceptual level. For example explaining the complexity of the phenomena that contribute to depression is now within grasp. I think people are finally beginning to appreciate that what is an apparently a unitary phenomenon in diagnostic classification systems may not be a unitary phenomenon at the level of biology. We are now beginning to fractionate out the relevant phenomena based around ideas such as how the brain reaps reward, how reward contributes to motivation, how we trade off between expectation of reward and punishment, issues I believe to be core to understanding the neurobiology of depression. I am confident the next 10 years will feed in to a richer and more grounded basis for psychiatric diagnosis.

What have developments in neuroimaging modalities such as fMRI and MEG revealed?

The importance of these neuroimaging modalities is that they enable us to understand how cognitive functions emerge not just from the activity of a single region of the brain but through engagement of a wide network of regions. Such a conceptualisation was not apparent from lesion deficit approaches. These techniques also enable us to understand brain plasticity at both a structural and functional level, quantify it, as well assess how pharmacological agents impact upon connectivity and plasticity. As a concrete example, my colleague Eleanor Maguire has shown that anterior hippocampus enlarges in Taxi drivers as they acquire navigational knowledge, and this is something that most neurologists would have dismissed as impossible 10 years ago. Imaging modalities, have their limitations but we are now seeing the development of methodologies that can overcome these limitations. A focus now is on having models of what constitutes the causes of the signals we measure using fMRI or MEG. One of my fellows, Rosalyn Moran, has developed a technique, a model based approach that enables us to extract information from an MEG signal regarding synaptic events, events below the spatial resolution of the technique but which can be accessed by having a specific model. This might sound like science fiction but it is no different than approaches adopted in say geophysics, meteorology or astrophysics. 

A large body of your research has examined human decision making, influenced by emotion…Could you give us an overview?

Decision making is a highly complex field. Much research rests on normative models in other words what is the optimal thing to do and the core idea is that a rational agent should maximise expected utility. However it turns out that in subjects do not all behave according to the prescriptions of normative theories. What seems to be the case is that there is no singular homuncular decision making system in the brain but multiple decision making systems. Some of these reflect mechanisms that emerged in the course of evolution and that, one assumes, enhanced adaptation to our environment. However the operation of these systems can impinge on the operation of other systems. In situations, where for example, there is uncertainty about an outcome, then a lower level system can generate an obligatory response, for example approach when the expectation is reward or avoid when the expectation is punishment, and these impulses can corrupt a rational decision making system. All of this might sound very abstract but in fact there are many every day examples. One of the best accounts of this is when we provide patients with information. How we frame that information, whether we tell a patient that the treatment is likely to result in a 85% probability of a cure or a 15% probability of no benefit will (although informationally similar) profoundly shape a patients subsequent choices.