The rapid contemporary development of wearable devices offers non-invasive and effective approaches for monitoring the human brain. Recent studies have investigated the prediction of epileptic seizures (ESs) using wearable measurements, such as scalp electroencephalography and functional near-infrared spectroscopy. The signal processing tasks are the core component of emerging closed-loop ES prediction (ESP) systems. Various research groups have introduced many state-of-the-art signal processing techniques to improve ESP performance. Wearable measurements consider low frequency and low spatial resolution characteristics. In this paper, we provide a comprehensive review of signal processing techniques including preprocessing, feature extraction, dimensionality reduction and classification schemes for ESP systems. Trends and concerns of ESP studies at the end of the manuscript.

当代可穿戴设备的迅速发展为监视人脑提供了非侵入性且有效的方法。最近的研究已经使用可穿戴的测量方法（例如头皮脑电图和功能性近红外光谱法）研究了癫痫发作的预测。信号处理任务是新兴的闭环ES预测（ESP）系统的核心组件。各个研究小组已经引入了许多最新的信号处理技术来改善ESP性能。可穿戴式测量考虑了低频和低空间分辨率特性。在本文中，我们对信号处理技术进行了全面的综述，包括ESP系统的预处理，特征提取，降维和分类方案。