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

理解人类大脑的关键挑战之一是弥合微观层面(神经元)和我们所知的人类有能力进行的丰富行为之间的“鸿沟”。除了是关于我们自己的最重要和最基本的问题之一，在多个尺度上了解大脑的结构和功能对于增加我们对神经和精神疾病(如癫痫、精神分裂症、抑郁症和阿尔茨海默氏症)哪里出了问题的了解至关重要。在大脑功能方面，我们知道信息在神经元产生的电信号中被表示和处理，信息通过白质纤维通路在大脑区域之间进行电子传输。目前，我们研究大脑功能最流行的成像技术是fMRI，它不能直接测量大脑的电活动，而是测量大脑区域在活动时氧合血液的增加。与这种间接测量相比，理想情况下，我们希望在执行各种认知任务时，非侵入性地检测大脑区域内和大脑区域之间流动的电活动模式。一种很有前途的技术是脑磁图(MEG)，它测量与神经元电流相关的弱磁场。它们透明地穿过头皮/头骨，然后可以用超导探测器阵列进行探测。除了是了解大脑电活动的直接窗口，脑磁图还可以以毫秒的时间分辨率测量活动，使我们能够跟踪电信号在大脑将不同区域的网络聚集在一起处理信息时迅速扫过大脑皮层的过程。这是功能磁共振成像根本做不到的。与传统的脑电电极相比，脑磁图也有一个优势，那就是它相对更容易找出大脑中活动的电源在哪里，因为它不会受到微弱电信号通过头骨和头皮泄漏到表面时发生的信息模糊的影响。脑磁图背后的基本技术始于20世纪70年代，但直到现在这项技术才真正成熟，我们才有了强大的全头多通道系统(200-300个传感器)。英国第一个这样的系统于2001年在阿斯顿大学安装，在过去的10年里，又开设了7个英国分校(约克、伦敦大学学院、卡迪夫、诺丁汉、格拉斯哥、牛津和剑桥)。然而，仍然存在一些挑战，特别是如果我们希望最好地利用MEG进行临床研究：2)这是一种新技术，有必要发展培训，以建立英国在这一领域的关键研究质量。2)脑磁图数据包含难以解释的神经信号的复杂混合，许多小组正在开发新的高级分析工具来尝试和解决这个问题-但仍然没有标准的分析方法。3)我们希望用来记录临床脑磁图研究数据的实验方案的标准很少。4)最强大的临床研究应用程序涉及对大量患者进行脑磁图扫描--一个网站本身很难做到这一点。出于所有这些原因，英国MEG的所有八个研究小组都希望走到一起，形成研究伙伴关系。该建议包括学术网络活动、培训计划、联合学生和建立统一的方法，以1)执行实验，2)分析脑磁图数据，3)为未来的大型合作项目存储数据。作为伙伴关系方案的一部分，我们还将从有限数量的参与者(每个地点80人)那里收集数据，以便我们能够试行建立脑磁图实验数据的共享数据库。如果我们希望对特定的临床人群进行大规模的协作研究，这是必不可少的。

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.