Presentation Information
[B-19-03]A Study on Continuous Blood Pressure Waveform Estimation from 77 GHz mm-Wave Sensor Signals Using Deep Learning
◎△Takahiro Ishimoto1, Takeshi Toda2 (1. Graduate School of Science & Technology, NIHON UNIVERSITY, 2. College of Science & Technology, NIHON UNIVERSITY)
Keywords:
blood pressure,mm-wave sensor,deep learning,ICEEMDAN,vital signs
Continuous blood pressure monitoring is crucial for extending healthy life expectancy and for the prevention and early detection of cardiovascular diseases. However, conventional cuff-based sphygmomanometers impose physical burdens and restrict measurement posture, making them unsuitable for long-term continuous monitoring. Therefore, this study investigates a method to estimate continuous blood pressure waveforms from millimeter-wave sensor signals. This is achieved by using a 77 GHz millimeter-wave sensor to measure minute displacements on the chest surface without physical contact, while utilizing continuous blood pressure waveforms acquired via a CNAP500 device as ground-truth signals. The method involves applying ICEEMDAN to the acquired sensor signals and inputting the extracted heartbeat-related components into a deep learning model to estimate the continuous blood pressure waveform.