Building multi-site clinical research capacity in Magnetoencephalography (MEG)

One of the key challenges in understanding the human brain is "bridging the gap" between the microscopic level (neurons) and the full richness of behaviour that we know humans are capable of. As well as being one of the most important and fundamental questions about ourselves, understanding brain structure and function at multiple scales is crucial for increasing our knowledge of what is going wrong in neurological and psychiatric diseases, such as Epilepsy, Schizophrenia, Depression and Alzheimer's.

In terms of brain function, we know that information is represented and processed in the electrical signals generated by neurons and that information is transmitted electrically between brain areas across white matter fibre pathways. Currently, the most popular imaging technique we have studying brain function is fMRI, which cannot measure the brain's electrical activity directly, but instead measures the increase in oxygenated blood that occurs in brain regions when they are active. Rather than this indirect measure, ideally we would like to non-invasively detect the patterns of electrical activity that flow within and between brain areas as we perform various cognitive tasks.

One promising technique is Magnetoencephalography (MEG), which measures the weak magnetic fields associated with neuronal electric currents. These pass transparently through the scalp/skull and can then be detected using an array of superconducting detectors. As well as being a direct window onto the brain's electrical activity, MEG can measure activity with millisecond time-resolution, allowing us to follow the rapid sweep of electrical signals across the cortex as the brain brings various networks of areas together to process information. This is something that fMRI simply cannot do. MEG also has an advantage over conventional EEG electrodes in that it is relatively easier to work out exactly where in the brain the electrical sources of activity are as it does not suffer from the smearing of information that occurs when weak electrical signals have to leak through the skull and scalp to the surface.

The basic technology behind MEG has been around since the 1970s, but it is only really now that the technology has matured so that we have robust whole-head multiple-channel systems (200-300 sensors). The first such system in the UK was installed at Aston University in 2001, and over the last 10 years another seven UK sites have opened (York, UCL, Cardiff, Nottingham, Glasgow, Oxford and Cambridge). However several challenges still remain, particularly if we wish to make best use of MEG for clinical research: 2) It is a novel technique and there is a need for developing training to build UK critical research mass in this area. 2) MEG data contains a complex mix of neural signals that are difficult to interpret and many of the groups are developing novel advanced analysis tools to try and solve this problem - but there is still no standard analysis approach. 3) There are few standards for the experimental protocols that we wish to use for recording clinical MEG research data. 4) The most powerful clinical research applications involve MEG scans on large numbers of patients - this is difficult for one site to do on its own.

For all of these reasons, all of the eight UK MEG research groups wish to come together to form a research partnership. The proposal consists of a mixture of academic networking activities, training programmes, joint studentships and establishment of unified approaches to 1) Performing experiments 2) Analysing MEG data 3) Storing data for future large-scale collaborative projects.

As part of the partnership programme we will also collect data from a limited number of participants (80 at each site) so that we can pilot the establishment of shared databases of MEG experimental data. This is essential if we wish to perform large collaborative studies on specific clinical populations.

The partnership proposal brings together all eight UK MEG research groups and includes a five-year programme of activities:

- We will arrange regular collaboration and management meetings between the eight sites. - Every year we will arrange training courses addressing a specific MEG research topic.

- We will offer 8 PhD studentships, recruited and starting as a single cohort but working on joint research programmes between at least two partner sites. Note that all of these studentships will be working on projects that address one or more of the MRC's priority areas for studentships (e.g. biomedical imaging, stratified medicine, systems medicine and interdisciplinary strategic skills)

- Each site will collect data using a set of standardised protocols over the course of one year. We propose to collect MEG data (and associated MR anatomicals for source localisation) from 80 individuals at each site. Note the intention is not to fully establish normative databases in each area, rather to act as a proof of principle that multiple sites, using different MEG scanners, can collect data using standardised protocols, assess the test-retest reliability of our measures and implement common analysis approaches.

- We will purchase a shared storage system, mirrored across four sites, for the data collected as part of this partnership.

- A bioinformatics research assistant will be appointed, for two years of the project, to implement the joint database and common approaches to data mining. Note that our ultimate goal is to establish a freely available 'virtual brain bank' of MEG responses to standardised protocols that any patient group across the world could be compared to. This partnership will allow us to take the initial steps towards this goal.

- We will instigate a MEG international travel awards scheme.

- We will offer travel/accommodation bursaries and invite international speakers to an annual "MEG UK" conference.

Impact on patient groups
----------------------------
Ultimately the collaborative work in this partnership will benefit patients with a wide range of neurological diseases, by increasing our ability and capacity to perform research on the underlying neurophysiology of conditions such as:

- Epilepsy: In the UK, multiple sites (Aston, Cardiff, UCL) currently host strong epilepsy programmes that will benefit from collaboration.
- Schizophrenia and other psychiatric disorders: Schizophrenia is a poorly understood pathology characterised by impairment in perception and expression of reality. Neural oscillations, which are integrally involved in short and long range connectivity, have been shown to be abnormal in schizophrenia and all of the UK sites are working on methods for connectivity measurement and applying those methods to patients.
- Developmental disorders: All the partner groups are developing MEG measured biomarkers of both healthy development and impaired cortical function and network connectivity in developmental disorders such as ADHD, autism and dyslexia.
- Diseases of old age: Similarly, many of the partner groups have MEG research projects studying healthy ageing and diseases of old age such as Alzheimer's (e.g. Cardiff, Cambridge, Oxford).
- Parkinson's Disease (PD): Neural oscillations in gamma, beta and <7Hz frequency ranges may be a biomarker of this disease. Multiple UK MEG sites are involved in attempting to further elucidate such biomarkers for PD subtypes (Glasgow, UCL, Nottingham).

Developing some understanding of their disorder is important to all patients and with this in mind, all of the partner sites will support regular discussion and dissemination of results from clinical studies to the local patient population. We will achieve this via regular accessible seminars, which will be held to inform patients (and interested members of the public) about results from our multi-centre studies. We will distribute results to healthcare professionals at each of the partner sites via separate seminars delivered by key clinical and basic science researchers.

Impact on the pharmaceutical industry
-------------------------------------------
The development of new neurological drugs is extremely expensive and all of the major companies are attempting to reduce costs by streamlining this process. A particular focus is the development of early biomarkers of drug action both within the patient group and in terms of predicting an individual's response. The proposed collaborative studies described here will demonstrate the utility of MEG (in some cases combined with other imaging modalities (e.g. magnetic resonance spectroscopy) to reveal the mechanism of action of drugs capable of crossing the blood brain barrier. We will maximize impact in this context by the following steps:

- Representatives from the pharmaceutical industry will be invited to attend both yearly conferences (MEG UK) as well as specialist meetings hosted by the sites leading these collaborative projects.
- Throughout the programme we will continue discussion with scientists from both the pharmaceutical industry and academic researchers with a view to preparing subsequent applications for funding under schemes such as the Developmental Clinical Studies program.
- We will organize a symposium on the use of MEG to assess the effects of pharmacological agents and cognitive training on brain networks, at MEG UK in 2014.

Outreach activities beyond patient populations
----------------------------------------------------
All of the partner sites have strong outreach programmes that are aimed towards making the results of scientific studies accessible to a broad non-academic audience. We will support outreach activities by making results of collaborative research available to each of the partner sites for use at such events. This will enable to public to see the clinical research that is currently underway in MEG.