心率
计算机科学
噪音(视频)
心电图
滤波器(信号处理)
心脏超声心动图
人口
人工智能
实时计算
模拟
算法
计算机视觉
血压
医学
心脏病学
内科学
环境卫生
图像(数学)
作者
Esteban J. Pino,Javier A. P. Chávez,Pablo Aqueveque
标识
DOI:10.1109/hic.2017.8227614
摘要
Ballistocardiogram (BCG) has been revisited in the last years as an unobtrusive method to detect heart beats. New electromechanical film (EMFi) sensors are now able to detect minimal oscillations in its surface, allowing to detect the mechanical action of the heart as it beats. This has allowed to develop unobtrusive systems for heart rate monitoring to be used as Point-of-Care devices, and to deploy them in waiting rooms, assisted living facilities or at home. In this work, an EMFi sensor is used to measure BCG via the pressure changes on the seat produced by the beating heart. In a lab environment, 34 healthy volunteers are measured under two conditions: at rest and after exercise, simultaneously with ECG. Also, in a clinical environment, 24 volunteers are also measured while waiting. The algorithm looks for the variability of the length transform at different scales or windows to determine a search window to detect beats from the BCG. A second correlation filter helps eliminate false peaks detected due to noise in the signal. Results show that in resting conditions, the mean error between the BCG HR and the reference ECG is only 0.4 beats per minute, with a standard deviation of 1.88. The noise rejection accuracy is 93%. The proposed algorithm can be used to identify beats and issue alarms under abnormal rhythms, providing timely alerts for at-risk population.
科研通智能强力驱动
Strongly Powered by AbleSci AI