An ultrasound distance measurement system for the Pulse Wave Velocity estimation

Cardiovascular diseases are a group of disorders of the hearth and blood vessels and are one of the worldwide main causes of death. In these years, the research focused on new techniques for improving the efficacy of preventive therapies. Currently, the arterial stiffness is widely used for cardiovascular risk stratification because it can lead to CVDs. The gold standard method for assessing the elastic properties of the arteries is the Pulse Wave Velocity (PWV) estimation. PWV is the speed of the blood pulse that propagates along the circulatory system. It is calculated by dividing the distance between the carotid and femoral sites by the Pulse Transit Time (PTT), that is the time taken by the wave to travel. The open problem is the lack of a standard method to measure this distance. Nowadays, the user takes the measurement with the tape meter, so the estimated distance is an approximation of the real path length. Another source of error depends on how much the user is precise in the measurement. For this reason, the aim of the thesis work is the implementation of a low-cost system that allows to measure distances with high accuracy in an automatic way, in order to replace the tape meter. For doing this, ultrasonic technology has been investigated. The system is composed of two sensors: the transmitter sends the mechanical wave and the microphone acquires the signal. The received signal is strongly related to the propagation delay of the ultrasonic wave. This last parameter is obtained by the implemented algorithms and, by knowing the speed of the ultrasonic wave, the distance can be obtained. Two approaches have been tested: the Pulsed method and the Frequency Modulated Continuous Wave (FMCW). In the first, the transmitter sends a sequence of bursts at its resonance frequency. The implemented algorithms are based on the Time-of-Flight method. Secondly, the Frequency Modulated Continuous Wave (FMCW) approach has been investigated, in which the sensor transmits the chirp signal. This method should provide higher accurate measurements. In this case, the time information is obtained from the phase difference between the two signals. Once the system has been developed, it has been tested with a set of distances. The measurement range is 30 – 100 cm because it is considered compliant with the aims of the PWV estimation. The size step is 2.5 cm, but in the range 50 – 75 cm it is 1 cm in order to evaluate the resolution of the system. For each approach, five acquisitions has been executed for each distance for evaluating the performance and repeatably of the system. The characterization of the system has been performed by measuring the linear distances, placing the two sensors face to face, and the angular ones. In this last case, the angle between the two probes varies for each distance. The obtained results show that the first method is more accurate, indeed the maximum error is less than 1 cm. The more reliable algorithm is the parabolic curve fitting, based on only two samples, on the envelope of the received signal. On the other hand, the second method shows higher errors. This is probably due to hardware limit of the sensors, as the frequency response. So, although this method is very efficient, it is highly affected by the performances of the selected sensors. In conclusion, the goals of the thesis work, thus the realization of a low-cost miniaturized system able to measure the distance in the range of interest, can be considered achieved.