Buttfield A C, Bolton M P
Clinical Engineering (Northwestern Medical Physics), Withington Hospital, Manchester M20 2LR, UK.
J Med Eng Technol. 2005 Jan-Feb;29(1):8-13. doi: 10.1080/0309190041000166234.
The accurate measurement of beat to beat intervals is essential for subsequent heart rate variability analysis. Where the ECG is used to derive the intervals, timing can be affected by artefacts such as muscle noise, electrode instability and also shape changes in the QRS complex. Identifying the QRS time by correlation methods can minimize the uncertainty but the method is computationally intensive. We have developed a real time RR interval measurement system using a correlation technique running on a low cost digital signal processor (TMS320C31). Sampling rate is 1 KHz. Timing resolution is +/- 1 ms. The correlation process uses an averaged complex from the actual ECG and has an adaptive noise threshold. The high processing speed of a DSP has proved ideal for accurate RR interval measurement. The system is described and test results with various signal to noise ratios and different types of noise are presented.
逐搏间期的精确测量对于后续的心率变异性分析至关重要。当使用心电图来推导间期时,计时可能会受到诸如肌肉噪声、电极不稳定以及QRS波群形状变化等伪迹的影响。通过相关方法识别QRS时间可以将不确定性降至最低,但该方法计算量很大。我们开发了一种实时RR间期测量系统,该系统使用在低成本数字信号处理器(TMS320C31)上运行的相关技术。采样率为1kHz。计时分辨率为±1ms。相关过程使用实际心电图的平均复合波,并具有自适应噪声阈值。事实证明,DSP的高处理速度非常适合精确的RR间期测量。本文描述了该系统,并给出了不同信噪比和不同类型噪声情况下的测试结果。