Design of a Barometer-Based Pulse-Taking Device With In Vivo Validation Against High-Frequency Ultrasound Pulse Wave Imaging

Wrist-site pulse-sensing devices have drawn increasing attention for daily monitoring of human body conditions in the context of western medicine and modernization of traditional Chinese medicine (TCM) palpation. Existing tactile pulse sensors are deemed to only access the superficial tissue layer as surface detectors but have not been validated against medical imaging of the radial artery and its pulsation. We hereby proposed to employ high-frequency (15.6MHz) and high frame-rate (1000fps) ultrasound radial pulse wave imaging (US-rPWI) to substantiate human radial pulse-taking. We first presented an easy-to-fabricate, cost-effective pulse-sensing prototype based on a commercial off-the-shelf (COTS) barometer. The prototype performance was evaluated using a homemade pressurized vessel-mimicking phantom and 30 different in vitro pulses outputted from a TCM palpation training machine. Phantom experiments evidenced accurate portrayal of dynamic pressure waveforms (mean distortion ratio: 3.38 +/- 0.07%, compared with the ground truth). In vivo validation on human pre- and post- prandial pulses was further performed. US-rPWI revealed significantly diminished pulse amplitudes and gradual waveform deformation from the arterial wall to the skin layer. Nevertheless, in vivo pulse waveforms detected at the skin surface by the prototype were highly correlated (correlation coefficient: 0.926 +/- 0.07) with those obtained from the radial arterial wall by US-rPWI. Moreover, the changes in post-prandial pulse waveforms detected by both the prototype and US-rPWI were in excellent agreement with those in TCM literature, thus demonstrating the sensitivity of human pulse-taking in quantitative assessment of health conditions at the point of care and providing more clinical evidence for the modernization of TCM sphygmology.

Automated summary

Wrist-site pulse-sensing devices have gained attention for monitoring human body conditions, but their accuracy has not been validated against medical imaging. This study proposed the use of high-frequency and high frame-rate ultrasound imaging to substantiate human pulse-taking. A prototype device was developed and validated in vitro and in vivo. The results showed a high correlation between the pulse waveforms detected by the prototype and those obtained from ultrasound imaging.