MRC Cognition and Brain Sciences Unit, University of Cambridge, UK; Department of Electrical and Computer Engineering, Tarbiat Modares University, Iran.
Department of Experimental Psychology, Complutense University of Madrid, Spain; Center for Biomedical Technology, Laboratory of Cognitive and Computational Neuroscience (UCM-UPM), Spain.
Neuroimage. 2022 Sep;258:119344. doi: 10.1016/j.neuroimage.2022.119344. Epub 2022 May 31.
Early detection of Alzheimer's Disease (AD) is vital to reduce the burden of dementia and for developing effective treatments. Neuroimaging can detect early brain changes, such as hippocampal atrophy in Mild Cognitive Impairment (MCI), a prodromal state of AD. However, selecting the most informative imaging features by machine-learning requires many cases. While large publically-available datasets of people with dementia or prodromal disease exist for Magnetic Resonance Imaging (MRI), comparable datasets are missing for Magnetoencephalography (MEG). MEG offers advantages in its millisecond resolution, revealing physiological changes in brain oscillations or connectivity before structural changes are evident with MRI. We introduce a MEG dataset with 324 individuals: patients with MCI and healthy controls. Their brain activity was recorded while resting with eyes closed, using a 306-channel MEG scanner at one of two sites (Madrid or Cambridge), enabling tests of generalization across sites. A T1-weighted MRI is provided to assist source localisation. The MEG and MRI data are formatted according to international BIDS standards and analysed freely on the DPUK platform (https://portal.dementiasplatform.uk/Apply). Here, we describe this dataset in detail, report some example (benchmark) analyses, and consider its limitations and future directions.
早期发现阿尔茨海默病（AD）对于减轻痴呆负担和开发有效治疗方法至关重要。神经影像学可以检测到早期的大脑变化，如轻度认知障碍（MCI）中的海马体萎缩，这是 AD 的前驱状态。然而，通过机器学习选择最具信息量的成像特征需要大量病例。虽然磁共振成像（MRI）有大量公开的痴呆或前驱疾病患者的公共数据集，但在脑磁图（MEG）方面却没有可比的数据集。MEG 在毫秒级分辨率方面具有优势，可以在 MRI 显示结构变化之前揭示脑振荡或连通性的生理变化。我们引入了一个包含 324 个人的 MEG 数据集：MCI 患者和健康对照者。当他们闭眼休息时，使用马德里或剑桥的两个地点之一的 306 通道 MEG 扫描仪记录他们的大脑活动，从而可以在不同地点进行推广测试。提供 T1 加权 MRI 以协助源定位。MEG 和 MRI 数据按照国际 BIDS 标准格式化，并在 DPUK 平台（https://portal.dementiasplatform.uk/Apply）上免费分析。在这里，我们详细描述了这个数据集，报告了一些示例（基准）分析，并考虑了它的局限性和未来方向。
