Abstract: The electrocardiogram(ECG) signals from wearable ECG monitoring equipment usually contain movement and device artifact.In this paper,a combined artifact recognition algorithm was developed based on the mutation degree of ECG amplitude and the connectivity of mutation distribution,the disorder degree of transformed ECG maximum and minimum values,and abnormal cardiac beat characteristics,etc.Embedding three key links in ECG automatic analysis,the combined artifact recognition algorithm was carried out,and selecting four kinds of data samples from three kinds of equipment,the algorithm was verified.Test results show that the combined artifact recognition algorithm can provide an artifact recognition sensitivity up to 98.35%,and improve QRS detection accuracy by 3.08%,at the same time,make the ECG automatic analysis operation time no increase and no rely on specific hardware equipment.The combined artifact recognition algorithm can be applied to ECG data analysis of various wearable devices universally and efficiently.