Fully-Customizable Wireless Wearable EEG Monitoring Platform with Motion Artifact Resilience and Input-Adaptive Active-Electrode Recording

脑电图（EEG）是一种测量人脑活动的非侵入性方法，具有高空间和时间分辨率。它的应用范围从监测和诊断各种神经系统疾病（如癫痫、阿尔茨海默病）到脑机接口。由于等待时间和医疗保健系统的负担（例如，设备、进行实验的训练有素的技术人员、进行分析的专家），在过去十年中开发了几种无线耳机，以实现动态（即门诊）脑电图监测和分析。这种可穿戴解决方案必须轻便、不显眼、无线、易于设置，并且具有合理的电池寿命。最重要的是，它应该能够以临床级的质量记录脑电图，同时消除运动伪影，这是无线脑电图耳机最常见的问题。我们开发了一种集成电路（IC），它采用了一种新颖的记录通道架构，能够实时去除运动伪影，同时对脑电图信号进行放大、采样和数字化，与现有技术相比（在功耗、信噪比等方面），其质量更高。与在后端信号处理单元中进行伪影检测/去除的替代解决方案不同，我们的解决方案中采用的通道内方法显着降低了记录电路所需的动态范围（因此，它的功耗，因此，更好的电池寿命），使每个记录通道成为一个独立的独立单元，无需中央处理单元，因此，最大限度地减少了创建多通道EEG记录设备所需的互连线的数量。这大大降低了设备的外形因素和复杂性。记录通道（即有源电极）的独立独立操作以及最小的互连线导致所呈现的技术在通道数量（即出色的可扩展性）及其放置方面完全可定制。

The electroencephalogram (EEG) is a non-invasive method for measuring a person's brain activity with high spatial and temporal resolution. Its applications span from monitoring and diagnosis of various neurological disorders (e.g., epilepsy, Alzheimer's disease) to brain-computer interfaces. Driven by the significant wait time and burden on the healthcare system (e.g., equipment, trained technicians for conducting the experiments, specialists for analysis), several wireless headsets have been developed over the past decade to enable ambulatory (i.e., outpatient) EEG monitoring and analysis. This wearable solution must be light-wait, unobtrusive, wireless, easy to set up, and with reasonable battery life. Most importantly, it should be capable of EEG recording with clinical-grade quality while removing motion artifacts, the most common problem with wireless EEG headsets.We have developed an integrated circuit (IC) that employs a novel recording channel architecture capable of real-time motion artifact removal while conducting amplification, sampling, and digitization of EEG signals with the superior quality compared to the state of the art (in terms of power consumption, signal-to-noise ratio, etc.). Unlike the alternative solutions where the artifact detection/removal is conducted in a backend signal processing unit, the in-channel approach employed in our solution significantly reduces the required dynamic range for the recording circuit (thus, its power consumption, hence, better battery life), making each recording channel a stand-alone independent unit, needless of a central processing unit, and consequently, minimizes the number of interconnecting wires required for creating a many-channel EEG recording device. This significantly reduces the device form factor and complexity. The stand-alone independent operation of recording channels (a.k.a., active electrodes) along with the minimal interconnecting wires results in the presented technology being fully customizable in terms of the number of channels (i.e., excellent scalability) and their placement.