Dr Christine Ecker
This profile is taken from the MRC Annual Review 10/11, Perspectives, which tells the stories of MRC scientists who made some of the most compelling research discoveries of 2010/11 by thinking about research problems from a new angle.
Lecturer in neuro-imaging at the Institute of Psychiatry (IoP), King’s College London.
Dr Christine Ecker has developed a new way of diagnosing autism from a 15-minute brain scan.
Christine is a neuroscientist who works on the MRC-funded Autism Imaging Multicentre Study (AIMS), a major research project exploring the relationship between genes and brain development in people with autism. AIMS is being carried out by the IoP and the universities of Oxford and Cambridge.
“Our team works with people with autism spectrum disorder and we’re mainly interested in brain structure and brain functioning; how the brain works in autism,” says Christine.
A difficult diagnosis
People with autism mainly have difficulty with social interaction and communication. It is a spectrum condition, so people can be affected by the condition to varying degrees and it is complex, involving many different interlinked factors. At the moment it is diagnosed on the basis of a person’s behaviour. But this is not always easy, as Christine explains:
“Diagnosis isn’t too complicated in children, because the defining symptoms start to appear around the age of four. But autism was only categorised as a disorder 10 or 15 years ago, so there are still a large number of undiagnosed adults out there, and diagnosing adults is a lot more problematic.
We give people tasks to do and interview their parents, but often that’s not too reliable because the people we see can be middle-aged, their parents might be in their seventies or eighties and you’ve got to ask them to remember how their son or daughter was at the age of four or five. So what we are trying to establish is a set of bio-markers for autism, to help with the diagnosis.”
A new scan for autism
In 2010, Christine and her colleagues began using a new analytical technique being developed at the IoP to look at autism. The technique they have developed involves doing a magnetic resonance imaging (MRI) scan of the person's brain and then ‘training’ a computer programme to analyse the differences in brain structure between people with brain disorders and healthy volunteers. The technique is extremely sensitive to all kinds of minute differences, such as volume, curvature and the pattern of folding in various parts of the brain. It puts these minute measurements together and then makes a prediction.
“The computer software will find a pattern of brain regions that are different between the two groups,” says Christine. “Once you’ve found this pattern you can scan someone new – and the computer programme will compare it to the sample it already has and then predict how likely it is that the person has autism.”
The team tested the technique on two sets of scans from around 40 patients from AIMS, half of whom had autism. In the first sample, which looked only at the volume of grey matter in the brain, the technique diagnosed autism with around 86 per cent accuracy. In the second sample, which measured all of the other features besides just grey matter volume, the accuracy was 90 per cent.
“What is new and exciting about our research is that for the very first time we can say something about individual brains. Traditionally we’ve only ever been able to look at mean differences between groups of people with autism and groups of healthy people. This could pave the way for a more individual approach to diagnosis and treatment”
Bringing the test to patients
Christine’s next challenge is to discover whether the scan will work in a clinical setting.
“We’re going to compare how the software performs in comparison to the conventional diagnosis. If it turns out that the test is still accurate then we can start with larger clinical trials. We're hoping to scan the next 100 people by the end of 2011 and move onto larger clinical trials in 2012,” she says.
“I don't think the brain scan will necessarily replace the conventional behavioural diagnosis, but it will add important extra information in cases where psychiatrists can’t be sure if someone has autism by using the conventional means.”
So if these studies eventually prove that the scan is an effective way of diagnosing autism, what will this mean for patients?
Christine says: “For most people, getting a diagnosis is very important because it unlocks access to specialist services. And of course it’s important to diagnose the condition as early and reliably as possible and particularly in adults where the diagnosis is very complicated and very time and cost intensive. So if we could diagnose autism with a 15-minute brain scan and patients don’t have to bring their parents along, it would make the diagnostic process itself a lot more comfortable for them.”
Watch a video of Christine speaking about her work
Published October 2011