研究目的
To propose a variational mode decomposition (VMD)-based heart rate estimation method using wrist-type PPG signals during physical exercise, focusing on removing motion artifacts (MA) for accurate HR monitoring.
研究成果
The proposed VMD-based HR estimation method effectively removes MA from wrist-type PPG signals during physical exercise, achieving satisfactory accuracy and robustness. It shows potential for use in wearable devices for health monitoring and fitness tracking.
研究不足
The method's performance may be affected by extremely strong MA during very complex motions, such as various forearm and upper arm exercises, running, jump, and push-up.
1：Experimental Design and Method Selection:
The study employs VMD for signal decomposition to eliminate MA, followed by a post-processing method to ensure robustness.
2：Sample Selection and Data Sources:
Utilizes PPG datasets from the 2015 IEEE Signal Processing Cup, including single-channel PPG signals, three-axis acceleration signals, and ECG signals.
3：List of Experimental Equipment and Materials:
Uses a pulse oximeter with green LED and a three-axis accelerometer embedded in a wristband.
4：Experimental Procedures and Operational Workflow:
Involves decomposing raw PPG signals using VMD, applying post-processing, and using acceleration data to further remove MA.
5：Data Analysis Methods:
Spectral peak tracking and verification algorithm is used to calculate HR from the cleansed PPG signal.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
