Evaluation of single-lead ECG signal quality with different states of motion

The mobile medical system with ECG is an important and widespread medical development. However, the dynamic ECG signal based on single-lead can create artifacts during acquisition. De-noising and feature extraction may compromise useful waveform characteristics, seriously affecting such subjects as follow-up identification and monitoring. Therefore, it is necessary to evaluate the quality of the dynamic ECG signal. A single-lead ECG signal quality evaluation algorithm is proposed in this paper. First, according to the three signal quality indexes of pSQI, kSQI and basSQI, a comprehensive dynamic ECG signal quality evaluation method based on the 0-1 programming concept was proposed for acquisition and de-noising. Second, hierarchical clustering based on the AGNES algorithm was adopted for QRS wave detection and single-cycle ECG signal extraction. Our experimental results show that our single-lead ECG signal quality evaluation algorithm can evaluate the quality of such parameters as ECG signal acquisition, de-noising, and QRS wave extraction and can determine whether the normalized single-cycle ECG signal can fully retain the original ECG signal characteristics. Our results also show that the algorithm is suitable for ECG quality assessment under different motion conditions.