In this paper, an ultra-low-power QRS-detection circuit is proposed which is suitable for wearable and implantable ECG-monitoring applications. Implementing the major part of the system in the analog-domain using current-mode circuits in the subthreshold region leads to an extremely low power consumption. The proposed circuit is based on calculating the energy of the derivative of the signal which follows by the onset spike detection method. Two low-pass filters with ultra-low cutoff frequencies of 6.45 Hz and 0.23 Hz are utilized in this method which are required for removing the high-frequency noise and estimating the required threshold for QRS-detection, respectively. The proposed algorithm is evaluated over the entire standard MIT-BIH arrhythmia database which results in a detection error rate of 1.46% with sensitivity and positive prediction of 99.17% and 99.36%, respectively. This QRS-detector circuit is implemented in 0.13 μm CMOS technology with a power supply of 1.2 V which consumes 31.5 nW power and occupies 0.097 mm² of silicon area. As the proposed circuit uses the analog ECG signal, there is no need to use an analog-to-digital converter which yields to further save in the area and power consumption of the entire system.

本文提出了一种适用于可穿戴和可植入ECG监测应用的超低功耗QRS检测电路。使用亚阈值区域中的电流模式电路在模拟域中实现系统的主要部分会导致极低的功耗。所提出的电路是基于计算信号的导数的能量的，该能量随后是峰值尖峰检测方法。此方法使用两个具有6.45 Hz和0.23 Hz的超低截止频率的低通滤波器，分别用于消除高频噪声和估计QRS检测所需的阈值。在整个标准MIT-BIH心律失常数据库中对提出的算法进行了评估，得出的检测错误率为1。敏感度为46％，阳性预测为99.17％和99.36％。该QRS检测器电路采用0.13μmCMOS技术实现，电源电压为1.2 V，消耗31.5 nW功率，占用0.097 mm硅面积²。由于拟议的电路使用模拟ECG信号，因此无需使用模数转换器，从而进一步节省了整个系统的面积和功耗。