Development of a lifespan compliant magnetoencephalography system

每200名儿童中就有1名患有癫痫。它是儿童时期最严重的长期健康状况之一，非常使人衰弱，影响身心健康。不幸的是，虽然可以使用药物治疗癫痫，但在约30%的病例中这些药物无效，许多患者在学习和行为方面存在困难。在精心挑选的儿童中，手术可以治愈，但这需要仔细规划，确保在不损害周围健康组织的情况下切除异常脑组织。规划需要先进的大脑成像技术，但现有的技术往往被证明是不够的。在儿童中，发生耐药癫痫的最常见原因是皮质发育异常，这种情况被称为局灶性皮质发育不良（FCD）。FCD有时可以在MRI扫描中看到，但它很微妙，经常被遗漏，因此需要其他技术来补充MRI。可以直接测量脑电活动，包括导致癫痫发作的脑电活动；要么是侵入性的（通过将电极插入大脑），要么是非侵入性的，通过头皮上的电极进行脑电图（EEG），要么是通过使用脑磁图（MEG）测量头皮上方的磁场。侵入性测量可以精确地定位癫痫发作的源头，但这需要大量的手术，而且只能评估大脑的一小部分区域（所以我们需要有一个明确的计划，在哪里放置电极）。脑电图在临床上广泛使用，覆盖了整个大脑，但它提供了癫痫发作的模糊图像。MEG提供了整个大脑活动的更详细的图像，并已被证明可以显著增加手术成功的机会。然而，目前的脑磁图扫描仪非常昂贵且不实用（因为患者必须长时间保持静止）。它们也不太适合儿童使用。最近，我们建造了一种新型的MEG扫描仪。与笨重的传统设备不同，我们的扫描仪可以像头盔一样戴在头上。因为扫描仪会随着头部移动，所以当病人做出较大的动作时，仍然可以产生扫描结果。此外，我们的可穿戴扫描仪可以更详细地测量大脑活动，而且更便宜，更容易维护。到目前为止，这种扫描器只为成人开发，我们现在计划为儿童设计和建造一个系统。我们必须解决一些主要的技术障碍：我们将从解决与扫描幼儿相关的基本问题开始，包括如何获得尽可能高的空间精度，以及如何确保磁场传感器（MEG系统的基本组成部分）在紧密地放在儿童头上时能够工作。我们将确保数据不受受试者运动的影响，我们将定制我们的阵列，专门关注已知易受FCD影响的大脑区域。我们将解决如何为婴儿制造可穿戴的脑磁图头盔的问题；让它既强大又实用，同时又能让孩子们（和他们的父母）乐于参与。我们将开发所需的数学方法，从脑磁图数据形成准确的大脑活动图像。最后，我们将在健康儿童（验证它）和癫痫婴儿中部署我们的系统。我们期望我们的系统将以无与伦比的准确性为神经科医生提供一个了解异常大脑功能的窗口。我们将把我们的结果与高性能MRI（以显示与FCD的一致性）和侵入性脑电图进行比较，表明我们的系统提供了与侵入性测量相似的信息，但不需要手术。最终，我们的目标是展示我们的设备提供了关于大脑异常功能的新信息，这将改变患有这种高度衰弱性疾病的年轻人的游戏规则。

Epilepsy affects around 1 in every 200 children. It is one of the most serious long-term health conditions in childhood and is highly debilitating, affecting physical and mental health. Unfortunately, whilst epilepsy can be treated using drugs, these are ineffective in ~30% of cases, and many patients experience difficulties in learning and behaviour. In carefully selected children, surgery can be curative, but this requires careful planning, ensuring abnormal brain tissue is removed without damaging healthy tissue surrounding it. Planning requires advanced brain imaging, but existing technologies often prove insufficient. In children, the most common cause of drug resistant epilepsy occurs is abnormal cortical development, a condition known as focal cortical dysplasia (FCD). FCD can sometimes be seen on MRI scans but it is subtle, and often missed, so other techniques are critically required to supplement MRI.It is possible to measure electrical brain activity, including that resulting in epileptic seizures, directly; either invasively (by putting electrodes into the brain) or non-invasively via electrodes on the scalp with electroencephalography (EEG) or by measuring magnetic fields above the scalp using magnetoencephalography (MEG). Invasive measures precisely pinpoint the source of the seizures, but they require significant surgery and only small regions of brain can be assessed (so we need to have a clear plan for where to put the electrodes). EEG is clinically widely available, covers the whole brain, but it provides a blurred picture of where seizures are generated. MEG offers a more detailed picture of activity across the whole brain and has been shown to significantly increase the chances of surgical success. However, current MEG scanners are extremely expensive and impractical (because patients have to keep still for long periods). They are also not well-suited for use in children.Recently, we have built a new type of MEG scanner. Unlike traditional devices which are large and heavy, our scanner can be worn on the head like a helmet. Because the scanner moves with the head, scans can still be generated when patients make large movements. In addition, our wearable scanner can measure brain activity with much greater detail and is cheaper and easier to maintain. Thus far, this scanner has only been developed for adults, we now plan to design and build a system for children. There are a number of major technical barriers that we have to address: We will start by tackling the fundamental problems associated with scanning young children, including questions like how to get the best possible spatial precision and how to ensure magnetic field sensors (the fundamental building block of a MEG system) can work when tightly packed together on a child's head. We will ensure that data are unaffected by subject movement, and we will tailor our array to specifically focus on brain regions known to be vulnerable to FCD. We will address the problem of how to actually build a wearable MEG helmet for infants; making it robust and practical, but also something with which children (and their parents) will happily engage. We will develop the mathematical methods required to form accurate images of brain activity from the MEG data. Finally, we will deploy our system in both healthy children (to validate it) and in infants with epilepsy. We expect that our system will offer neurologists a window on abnormal brain function with unparalleled accuracy. We will compare our results to high performance MRI (to show concordance with FCD) and with invasive EEG, showing that our system offers similar information to invasive measurements, but without the need for surgery. Ultimately, we aim to show that our device offers new information on abnormal brain function which will be game-changing for youngsters suffering with this highly debilitating disorder.