There are a number of different ways to get a picture of brain activity, all of which are flawed to some degree. Positron emission tomography, for example, allows us to see deeper into the brain than many other techniques. But it exposes the patient to high levels of ionising radiation that, while not thought to be dangerous, are best avoided if possible.
The most durable method has proven to be functional magnetic resonance imaging (fMRI): a non-invasive technique that has been around since the early 1990s, and is now the most popular way of measuring and mapping the human brain. Though it can be used in a number of ways, the most common form of fMRI is known as blood-oxygen-level-dependent (BOLD) contrast imaging, a technique pioneered by Seiji Ogawa – an AT&T Bell Laboratories scientist and the man generally regarded as the father of functional brain imaging.
In layman’s terms, BOLD imaging measures brain activity by monitoring changes in blood flow. Oxygen-rich and oxygen-poor bloods have different magnetic properties (due to the iron levels contained therein) and when a part of the brain is more active, it receives greater oxygen-rich bloodflow. When oxygen-rich blood flows in this way, its contrasting magnetic state is registered on the fMRI scan, creating a multicoloured 3D image of brain activity. Such scans give a good idea of how certain stimuli affect activity in different parts of the brain, making them a valuable tool for medical researchers and clinical psychologists.
Employed in the field
Increasingly, fMRI and the BOLD imaging technique are being used for more functional, clinical purposes. Resting-state fMRI data (gathered when the brain is not performing a specific task) is being analysed to give greater insight into the development of diseases. Spontaneous BOLD fluctuations measured in the brains of patients at rest have revealed that healthy subjects share a number of what researchers at the University of Lund refer to as “spatially consistent networks”. Being able to track the evolution of these networks, and how they are affected by aging and the onset of disease could lead to a far deeper understanding of the mechanics of brain disease.
Although most of this research is being done in North America and Europe, clinicians in the Middle East are keeping a close watch on developments. In July last year, Dr Mwahib Sid Ahmed Mohammed Osman Aldosh, assistant professor and coordinator of the radiological sciences department at the Applied Medical College at Saudi Arabia’s Najran University, published a paper entitled ‘Role of (fMRI) in Clinical Applications in the Field of Neurosurgery’ in The International Journal of Science and Research. The paper provides an interesting overview of the current usefulness of fMRI in brain surgery and of some of the newer techniques being employed.
The crux of his article is that the study of these ‘spatially consistent networks’ with fMRI is already having an impact on the way surgeons approach their jobs. For example, fMRI monitoring increasingly allows them to identify functionally important parts of the brain that lie in the vicinity of a tumour or a piece of the brain that is set to be removed, ensuring that nothing vital is damaged during the extraction process.
Although the structure of human brains tends to be more or less the same in each case, there can be subtle but important differences that fMRI is able to flag up – vital functional parts that appear in unexpected places. For example, sometimes if a person suffers brain damage early in life, other parts of the brain can take over the lost functions, changing the physical makeup of the brain. It has also been known for stroke victims to regain some lost functionality in the corresponding part of the unaffected hemisphere of their brain.
“Brain mapping helps determine the way a specific patient’s brain is structurally and functionally arranged,” Aldosh says. “Although the gross anatomical structure and functions of the brain are quite similar in all humans, the detailed organisation can vary amongst individuals. For example, portions of the language areas are quite variable between individuals, while motor areas are very similar… fMRI can be used to predict the functional outcome of patients suffering from trauma, stroke, schizophrenia and Alzheimer’s disease.”
A large part of why MRI is very much the favoured tool for diagnosing primary and metastatic tumours is the clear contrast it shows between different areas of soft tissue, but its ability to identify tumours has improved a lot in the past several years. Contrast agents made of gadolinium, a rare earth element, have helped bring out tumours in areas where they were once difficult to spot – most notably the blood-brain barrier (the semipermeable membrane that separates blood from the cerebrospinal fluid). Three-dimensional imaging is also being used pre-operation to figure out how best to attack the tumours – generally stereotactic radio surgery for a smaller tumour and neurosurgical resection for a larger one.
“This conventional MRI can be used for planning the best way to surgically remove or destroy intracranial masses,” says Aldosh: “fMRI can provide the functional information about nearby tissue that may be affected by neurosurgery or stereotactic radiosurgery. It also helps predict the clinical outcome if nearby tissue is inadvertently damaged. The surgical approach to a lesion can be better planned because the surgeon is aware of the location of important functional areas, such as a motor or language centre.”
Some of the most interesting developments are being seen in the treatment of epilepsy, where fMRI is being effectively combined with another visualising tool: electroencephalography (EEG). EEG involves small sensors attached to the scalp that pick up electrical signals when the brain cells fire messages to each other. Often the EEG will pick up electrical spikes or waves of spikes caused by hyperactive neurons in the brain; these don’t have any external effect on the patient but are very important, particularly in predicting whether epilepsy will develop in the aftermath of a stroke.
Unfortunately, EEG has very low spatial resolution; it can tell that a neutron is firing but is not able to identify the precise location of that neutron. By using EEG and fMRI at the same time, clinicians can identify these spikes with the former and more accurately pinpoint their location with the latter.
“fMRI combined with EEG is a new diagnostic tool in epilepsy and sleep disorders,” Aldosh says. “fMRI can identify the penumbra after stroke [the most common cause of epileptic seizures among older people] and can provide additional information on the metabolic state of the threatened brain tissue. It can have a predictive role in post-stroke recovery.”
A long way to go
Despite the great progress being made in brain imaging, there are still a number of obstacles to be overcome before fMRI can take a bigger role in clinical life, many of these related to the technology itself.
fMRI measurements are often spoiled by noise from other sources: thermal, system and physiological noise, and random neural activity, among other types. The underlying signal must be extracted from the din through a series of complex statistical techniques. While these techniques are becoming more effective, when analysing the tens and thousands of complex, ever-shifting data points found in the human brain, it is easy to see coincidences in how the brain reacts to certain stimuli and mistake them for patterns.
Neuroscientists such as Chris Chambers from Cardiff University also argue that studies done so far have been of insufficient statistical power to reveal any concrete conclusions.
“Evidence from structural brain imaging implies that most fMRI studies have insufficient sample sizes to detect meaningful effects,” Chambers wrote on the Neurochambers blog. “This means they not only have little chance of detecting true positives, but there is also a high probability that any statistically significant differences are false.”
This view is shared by senior researchers Johan Olsrud and Peter Mannfolk at the University of Lund, in the abstract of a paper entitled ‘Optimisation of clinical fMRI’. They argue that it will be some time before fMRI becomes an integral part of the neurosurgical offering. Meanwhile, many of the new avenues of fMRI research look not at blood oxygen levels, like BOLD, but at other biomarkers such as temperature and PH level, which are less prone to disruption from noise.
“Although there is much enthusiasm,” Olsrud and Mannfolk write, “few clinical applications of fMRI exist, and the ones that are emerging are still prone to many sources of error and are not trivial to implement. Depending on the application, different issues should be addressed, including image acquisition, pre-processing and statistical analysis of data. A development of all these steps with a view to more robust results, even in individual examinations, will aid in the development of fMRI as a useful clinical tool.”
