Trevor Robbins: Can AI detect mental health disorders?

In this edition of Titans of Science, Chris Smith chats to cambridge Neuroscientist, and expert on OCD, Trevor Robbins...

Chris - When you were doing these sorts of studies, you're using MRI scans. You put a person in a scanner and you're interrogating which bits of the brain are talking to each other. Often it is painstaking and it takes absolutely ages because you're looking for very subtle signals. We've now got tools like AI, haven't we? Are you beginning to look at those images now through the lens of AI to ask it to spot some of the more subtle things we might have missed?

Trevor - I think this is definitely the future. You can get so far with the rather sophisticated analyses that have been done already. However, putting two and two together often requires something as clever as AI to notice that a change here and a change there are actually related to one another and part of the same general circuit. So I think that's definitely something in the future.

Chris - Since presumably if we can see what circuits are doing to each other, we might then have some drugs which are discreet for those circuits and which, which might help to rebalance that, that imbalance that you've spotted that little bit better and with fewer side effects.

Trevor - Yeah, I mean that would be the great aim and we've certainly ourselves been developing this idea using a class of drugs which may reduce glutamate for example, in those areas where it's more active. I mean that would be an interesting approach. But there's a general problem with drugs, as you know, that the receptors for these drugs are often all over the brain and so that's what leads to the side effects. You know, you might be normalising something in one area but overdosing or whatever another area, and that is problematic. So how do we get that specificity? You can't necessarily rely on where the receptors are. Maybe you can get lucky about that. In the future, we may need to use techniques such as chemogenetics. Chemogenetics of course are only used in animals at the moment, but with this technique you can take a pill which activates or inhibits a particular neural circuit in the brain.

Chris - I've not heard of this. How far down the track is that?

Trevor - As I said, it's only been used in experimental animals so far, like mice. But what one has to do is to put a virus in the brain, a benign virus, that is done actually already in Parkinson's disease in some cases. And that puts a receptor in the brain, which has been engineered so that it doesn't respond to anything normally, but you put it in a particular circuit and then when you take a pill, a pill that's designed specifically to activate that receptor, you can achieve an excitation or an inhibition of that circuit. And it's the way of really analysing in minute detail, which circuits are in control of different behaviours. We are not going to be able to do that in humans <laugh> for quite a while yet because potentially it involves surgery. But it's a very interesting approach, particularly for severe patients. It's obviously more specific than something like deep brain stimulation, for example,

Chris - Any notable figures back in history who were affected? Because when one looks back in the history books, once we see what we think is a modern phenomenon, we then begin to see evidence that perhaps other people had this back in the past. Anyone stand out?

Trevor - Martin Luther, the Reformation, he was an obvious OCD patient, if you look at his biography. Kurt Gödel the famous mathematician. Howard Hughes, of course the great filmmaker, who had this problem with germs. His entire staff spent all of their time cleaning the house and cleaning his food and so forth. Charles Dickens had a bit of OCD. Samuel Johnson, Andy Warhol, Cameron Diaz, the actress. So it's interesting that in many of these cases people have been able to cope with this problem and be pretty effective in life, but many of them really are on the cusp of OCD and didn't really have a full blown diagnosis. Another example is David Beckham.

Chris - It slightly troubles me that if we were to go down the path of offering people help, we might have not had Charles Dickens and we might not have been able to bend it like Beckham.

Trevor - Well, indeed it becomes an issue about when these behaviours become excessive, but it's up to really the people and their relatives to understand that and then to get some help.

Chris - And in the years ahead. Now looking back at where you've come from and where we've now got to, what are the questions that you would really like to, in the years you have left on Earth, see answered in the OCD and frontal lobe space?

Trevor - What I really want to do is to achieve effective treatments. I wouldn't say cures, but you know, near cures for OCD. That would not only give me a lot of satisfaction to help people, but it would also serve to confirm some of the theories that I have about how the brain works. And I think that dual motivation is very important to me. I'm very interested in how the frontal lobes control behaviour, how neurochemical systems modulate this function and how this develops. And also its relationship to philosophical issues. For example, free will, which is clearly very important here.