The energy-efficient implementation of an adaptive-filtering-based QRS complex detection method for wearable devices

Electrocardiogram (ECG) can be used as a valid way for diagnosing heart disease.To fulfill ECG processing in wearable devices by reducing computation complexity and hardware cost,two kinds of adaptive filters are designed to perform QRS complex detection and motion artifacts removal,respectively.The proposed design achieves a sensitivity of 99.49％ and a positive predictivity of 99.72％,tested under the MIT-BIH ECG database.The proposed design is synthesized under the SMIC 65-nm CMOS technology and verified by post-synthesis simulation.Experimental results show that the power consumption and area cost of this design are of 160μW and 1.09 × 105 μm2,respectively.