一种非接触式人体生理信号监测方法

摘要

Non-contact biosignals monitoring system based on mattress type uses two-order least mean square (LMS) adaptive filtering to extract effective respiration and heartbeat signal, can realize dynamical biosignals extraction in unlimited breath and heartbeat filtering frequency band range, solves problem that abnormal heartbeat signal is difficult to separate from respiration signal fixed filtering frequency band. Experiment with 20 groups of respiration and heartbeat seperation result in gentle respiration and rapid respiration situation compared with multi-parameter monitors shows that the accuracy of this algorithm is up to 90 % and has good practicality.%基于床垫式非接触人体生理信号监测系统,采用二阶最小均方(LMS)自适应滤波提取有效的呼吸和心跳信号,可实现不限定呼吸和心跳滤波频带范围的动态生理信号提取,解决了固定滤波频带下异常心跳和呼吸信号不易分离问题.在平缓呼吸和急促呼吸2种情况下,20组呼吸和心跳分离结果与多参数监护仪结果的对比分析表明:该算法的准确率可达90％以上,具有较好的实用性.