基于三轴加速度信号的心率修正算法

HEART RATE CORRECTION ALGORITHM BASED ON THREE AXIS ACCELERATION SIGNAL

目前大多数腕带式心率监测设备获得心率的算法主要都是基于光电容积脉搏波描记法PPG(Photo Plethysmo Graphy)。但是在剧烈运动情况下仅依靠PPG信号获得心率的算法准确度大大降低,并且由于运动伪影MA(Motion Artifacts)的存在大大增加了PPG信号进行频谱分析的难度。为了进一步修正心率检测算法,通过对受测者运动分析,提出一种使用三轴加速度计信号进行修正心率的算法,由运动强度估算出的心率和将PPG信号得到的心率与修正心率通过卡尔曼滤波器处理得到最终心率两部分组成。实验数据记录了由10位受测者以10~16 km/h进行快速跑步的结果,测试数据表明,提出的心率修正算法的绝对平均误差为2.05 BPM。

Recently, heart rate monitoring using wrist-type PPG （PhotoPlethysmoGraphy） is the main measurement device for the heart rate detection. But it is a difficult problem, since the signals during subjects＇ intensive exercise are contaminated by strong motion artifacts caused by subjects＇ hand movements. The accuracy of the algorithm based on PPG signal to obtain heart rate is greatly reduced, and the existence of motion artifact increases the difficulty of spectrum analysis of PPG signals. In this study, we propose a modified algorithm for heart rate monitoring based on three-axis accelerometric data, it consists of two key parts, and heart rate estimated by exercise intensity and obtained final heart rate by caiman filter. Experimental data recorded the fast running results of ten subjects with 10 - 16 km/h, and the test data showed the absolute mean error of the proposed heart rate correction algorithm was 2.05 BPM.