Dick Veltman: Scanning for trouble in the brain

Brain imaging has brought to psychiatry the possibility to directly observe abnormalities in brain function. Having worked on the visualisation of anxiety and depression for over a decade, Professor Dick Veltman expects imaging techniques to impact on clinical practice in the years to come.

'When I wrote my thesis on anxiety disorders in 1995, I already focused on both the cognitive psychology and the biological changes that are associated with the disorder. But we did not have the scanning machines that we have today, so for the biological aspects I had to work with peripheral measures such as blood pressure and heart rate. When the VU stepped up its PET imaging activities, I readily volunteered to work on that subject, in close collaboration with professor Lammertsma. Together, we gradually set up a line of research in neuroimaging. In the beginning the project was confined to the VU, but later on we extended our collaboration with the psychiatry department of the AMC, for which I now also work.'

Much of the work still consists of the development and validation of imaging methods, the professor of neuroimaging explains. The scanning techniques are getting better: the resolution is improving, more specific labelling and contrast substances are becoming available and new data processing algorithms are being developed. 'Functional MRI still is a fairly noisy technique. The noise is often much louder than the signal that we are interested in. In a different way though, this also holds true for MEG and PET scanning. Separating the signal from the background activity is a delicate process: if we are too lenient, we would end up analysing nonsense signals, while being too strict would erase everything, including the activity we are studying.'

Initially, much of the brain imaging work yielded only anatomical data. 'In the beginning we tried to pinpoint specific cognitive processes to a specific location in the brain, mainly using healthy controls in functional MRI studies. Now that we know where certain brain functions can be found, we can also compare healthy controls with patients. For schizophrenia we know for example that short-term memory is impaired, so we can investigate the differences between a patient and a control subject during a memory task. At this moment we know quite a bit about the function and malfunction of biological processes in the brain of psychiatric patients.'

Students using XTC
The field is now moving towards intervention studies. At first, patients are compared to control subjects by for example functional MRI. Then the intervention takes place, for example a treatment with an anti-depressant, and the result of the treatment on brain function is measured by another scan. But the intervention is not necessarily therapeutical, Veltman explains: 'We also focus on groups with an enhanced environmental or genetic risk for psychiatric disease. When we find functional differences between addicted individuals and normal controls, we would like to know if these differences are a result of the habit or if they predate it. Researchers for example could scan students in their first year, and then again in the third year. Some of them will start to use XTC, a voluntary intervention, which allows the investigators to observe the causality. Another advantage of this type of study is that the variance is lower when you scan the same person multiple times.'

Veltman is convinced that brain imaging will become a regular instrument in the care for psychiatric patients. 'In the end we want to use the techniques to improve the treatment of individual patients. Our approach is not fundamentally different from, for example, measuring the content of haemoglobin in the blood: we too need a big control group to relate the outcomes of such a functional brain image to. Before we can use these techniques as a diagnostic tool, we need to further improve their sensitivity and specificity. But we already know that certain abnormalities in functional scans can predict whether a patient is going to respond to a certain antidepressant. And later on, we may be able to use scans to monitor the effectiveness of therapy.'