I lay flat on my back on the table. Large foam pads were placed under my elbows and round my neck. I was given a pair of headphones and a restraining strap was fixed across my forehead. Then I was shunted backwards into a narrow, cream-coloured tube. The door was locked and the room was filled with an incessant humming noise as the machine started. Soon the humming was replaced by squealing and the whole machine began to vibrate.
I was volunteering for an MRI (Magnetic Resonance Imaging) brain scan and whilst I was in the scanner I had to perform various tasks, including listening to noises and repeating nonsense words. The scan was part of an ongoing experiment on language processing being conducted by Gayaneh Szenkovits, a member of the research staff at the Medical Research Council Cognition and Brain Sciences Unit in Cambridge.
Being a volunteer requires a level of time and commitment; when experiments are conducted across multiple sessions, volunteers are required to come back for follow-up sessions otherwise researchers cannot make good use of the collected data. Therefore, all MRI scan volunteers in Cambridge are paid ten pounds an hour by the MRC. After my scan, Gayaneh tells me that participants occasionally fall asleep in the scanner – most humans simply aren’t used to lying down for long periods of time without falling asleep. And all the padding round my head and neck was to stop me moving. Corrections can be made for small motions, but too much fidgeting renders the entire scan useless.
MRI scans have become increasingly popular amongst neuroscientists in recent years, providing a low-risk, non-invasive method of data collection. Structural scans are used to look at brain morphology, whilst functional scans are able to highlight which areas of the brain are active when participants perform certain tasks. Functional MRI actually monitors blood flow in different portions of the brain by comparing the concentrations of oxygenated and deoxygenated haemoglobin. However, the exact form of the correlation between blood flow and neural activity is still an active topic of research. Another issue is that a correlation between a certain mental process and increased blood flow to a given portion of the brain doesn’t necessarily imply a causal link. Furthermore, data must be analysed carefully: researchers once reported brain activity in a dead salmon simply by applying the wrong statistical analysis to a dataset!
Dr Gayaneh Szenkovits is using MRI to investigate phonological processing, that is, the processing of the sounds of language, which is thought to take place in the brain’s left hemisphere. Syllables are formed of segments (or phonemes), and these are composed of certain ‘distinctive features’. But what is it about speech that sets it apart from other auditory information? How do our brains decode meaning from the sounds?
Dr Szenkovits’ research focuses on the differences between dyslexics and non-dyslexics. It is likely that phonological processing ability lies on a continuum. However, she is only able to scan fifty participants, so she is using two discrete groups of participants from either end of the spectrum. She tells me that she expects neural activity to be weaker and more dispersed in dyslexics as compared to non-dyslexics. She is also finding that dyslexics appear to make use of complex compensatory mechanisms to help with their deficient phonological processing, possibly relying on their brain’s right frontal lobe.
Dr Szenkovits’ research is fascinating, especially the way she has to design her experiments to target the very specific mental processes she is interested in. But it is equally astonishing how little we know of the brain’s inner workings.
Further research relies heavily on volunteers to take part in experiments. I very much enjoyed being involved but nevertheless, the experience was a strange one: simultaneously claustrophobic, soporific and surprisingly draining.