.png){kind=icon}
.png){kind=icon}
Return
Pulse wave signal acquisition and processing using flexible pressure sensors for cardiovascular disease monitoring
.png){kind=icon}
PRE
AI
X
X
Y
Z
J
C
L
DOI：10.1016/j.bspc.2026.111189.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Cardiovascular diseases (CVDs) are the leading cause of death worldwide, and early monitoring is essential for health management. This paper proposes a CVD monitoring method based on a PDMS/GNPs piezoresistive flexible pressure sensor. Sensors with varying parameters were fabricated, and testing results demonstrated that the sensor with 8 wt% filler mass fraction and 1.0 mm thickness exhibited optimal performance, achieving a sensitivity of 0.17 kPa−1, a response time of approximately 80 ms as well as excellent flexibility and durability. Subsequent signal processing was implemented to improve classification accuracy. Firstly, raw pulse wave signals were denoised using wavelet transform combined with cubic spline interpolation, followed by signal normalization and periodic averaging for waveform stabilization. Secondly, a one-dimensional convolutional neural network (1D-CNN) was adopted to detect critical feature points and extract multi-dimensional features in the time, frequency, and time–frequency domains. Finally, a Support Vector Machine (SVM) was applied to classify CVDs. Experimental results indicated that this classification model achieved 90.81% accuracy in distinguishing pathological pulse signals from normal ones. This study provides a promising approach for intelligent identification in wearable CVD monitoring applications.
Journal
IF:
4.9
Papers:
9.7K
Citations:
2.4W
Organization
Cited Papers