We show that a non-intrusive pneumatic sensor can be used to measure respiratory rate, heart rate, and their variability during sleep. The pneumatic sensor was included in a polysomnography (PSG) study involving 42 participants in a sleep laboratory. The agreement between the pneumatic sensor and the PSG for respiratory rate, heart rate, and their variability was quantified by Bland-Altman analysis. The respiratory rate has a mean value of 15.4 breaths per minute for a bias of −0.06 and 95% limits of agreement (LOA) of [−0.6; 0.5] breath per minute. The respiratory rate variability root mean square of successive differences (RMSSD) has a mean value of 459.51 ms, a bias of 9.2 and a 95% LOA of [−43.5; 61.9] ms. The heart rate has a mean value of 60.6 beats per minute for a bias of −0.8 and a 95% LOA of [−4.3; 2.7] beat per minute. The heart rate variability RMSSD has a mean value of 44.1 ms, a bias of 14.7 and a 95% LOA of [−19; 48.4] ms. These results show that a non-intrusive pneumatic sensor can accurately estimate cardiorespiratory metrics overnight.

Session 4aEAa: Non-Contact Ultrasonic Methods 4aEAa7. Coupling of finite difference elastodynamic and semi-analytic Rayleigh integral codes for the modelling of ultrasound propagation at the hip

International audienceThe development of novel quantitative ultrasound (QUS) techniques to measure the hip is critically dependent on the possibility to simulate the ultrasound propagation. One specificity of hip QUS is that ultrasounds propagate through a large thickness of soft tissue, which can be modeled by a homogeneous fluid in a first approach. Finite difference time domain (FDTD) algorithms have been widely used to simulate QUS measurements but they are not adapted to simulate ultrasonic propagation over long distances in homogeneous media. In this paper, an hybrid numerical method is presented to simulate hip QUS measurements. A two-dimensional FDTD simulation in the vicinity of the bone is coupled to the semi-analytic calculation of the Rayleigh integral to compute the wave propagation between the probe and the bone. The method is used to simulate a setup dedicated to the measurement of circumferential guided waves in the cortical compartment of the femoral neck. The proposed approach is validated by comparison with a full FDTD simulation and with an experiment on a bone phantom. For a realistic QUS configuration, the computation time is estimated to be sixty times less with the hybrid method than with a full FDTD approach

No abstract