Kappel Simon L, Makeig Scott, Kidmose Preben
Neurotechnology Lab, Department of Engineering, Aarhus University, Aarhus, Denmark.
Department of Electronic and Telecommunication Engineering, University of Moratuwa, Katubedda, Sri Lanka.
Front Neurosci. 2019 Sep 10;13:943. doi: 10.3389/fnins.2019.00943. eCollection 2019.
Computational models for mapping electrical sources in the brain to potentials on the scalp have been widely explored. However, current models do not describe the external ear anatomy well, and is therefore not suitable for ear-EEG recordings. Here we present an extension to existing computational models, by incorporating an improved description of the external ear anatomy based on 3D scanned impressions of the ears. The result is a method to compute an ear-EEG forward model, which enables mapping of sources in the brain to potentials in the ear. To validate the method, individualized ear-EEG forward models were computed for four subjects, and ear-EEG and scalp EEG were recorded concurrently from the subjects in a study comprising both auditory and visual stimuli. The EEG recordings were analyzed with independent component analysis (ICA) and using the individualized ear-EEG forward models, single dipole fitting was performed for each independent component (IC). A subset of ICs were selected, based on how well they were modeled by a single dipole in the brain volume. The correlation between the topographic IC map and the topographic map predicted by the forward model, was computed for each IC. Generally, the correlation was high in the ear closest to the dipole location, showing that the ear-EEG forward models provided a good model to predict ear potentials. In addition, we demonstrated that the developed forward models can be used to explore the sensitivity to brain sources for different ear-EEG electrode configurations. We consider the proposed method to be an important step forward in the characterization and utilization of ear-EEG.
用于将大脑中的电信号源映射到头皮电位的计算模型已得到广泛探索。然而，当前模型对耳廓解剖结构的描述不佳，因此不适用于耳脑电图记录。在此，我们通过纳入基于耳朵三维扫描印记的改进耳廓解剖结构描述，对现有计算模型进行了扩展。结果得到一种计算耳脑电图正向模型的方法，该方法能够将大脑中的信号源映射到耳朵中的电位。为验证该方法，为四名受试者计算了个性化的耳脑电图正向模型，并在一项包含听觉和视觉刺激的研究中，同时记录了这些受试者的耳脑电图和头皮脑电图。使用独立成分分析（ICA）对脑电图记录进行分析，并利用个性化的耳脑电图正向模型，对每个独立成分（IC）进行单偶极子拟合。根据它们在脑容积中由单个偶极子建模的效果，选择了一部分IC。计算了每个IC的地形图IC图与正向模型预测的地形图之间的相关性。一般来说，在最靠近偶极子位置的耳朵中相关性较高，这表明耳脑电图正向模型为预测耳朵电位提供了一个良好的模型。此外，我们证明了所开发的正向模型可用于探索不同耳脑电图电极配置对脑信号源的敏感性。我们认为所提出的方法是在耳脑电图表征和利用方面向前迈出的重要一步。
