Development of a cuff-less Blood monitoring device

In an era where technology is advancing at an unprecedented rate, surpassing itself day by day, it is inevitable that this progress involves the health area too. This evolution is mainly represented by the constant increase in user-friendliness, functionality, miniaturization of devices and the possibility of collect a wide amount of data about a subject.In these devices the increment of computational power and the easiness of the integration with complex algorithms encourage their application on health technologies. The study of new methods for evaluating different vital signs falls within this scope.The objective of this thesis work consists in the use of a wearable device for evaluation of Blood Pressure(BP) in a non-invasive cuff-less based way. In fact, BP is one of the alerts for cardiovascular disease(CVD). In the last few years, some of major causes of death worldwide are cardiovascular disease (CVD). The negative trend of unhealthy lifestyles, together with the rising age in today’s world, is closely related to the rising of number of people suffering of CVD and, one of the risk factors, is chronic hypertension, characterized by high blood pressure retained for a long period of time.Therefore, the continuous monitoring of blood pressure (BP) allows to provide a valid tool for observing patients in significant health conditions (e.g. for those who have CVD such as hypertension avoiding a degeneration into heart attack and stroke) but it is also an important instrument of diagnosis.Wearable Health Devices are every year more present in daily use, progressively helping people to monitor their health condition both at a sporty level, for optimization of sport activity, and at a medical level, with a monitoring and prevention value. The rise in the fabrication and use of wearable devices in daily life, along with the development of their reliability and precision, promote the perspectives of development of this market in the coming years. Blood pressure measurement enters in this field have been explored physiological parameters that allow to measure blood pressure in a roundabout way. The main one is the Pulse Wave Velocity(PWV) that encloses most of connections with BP. In PWV values are contained the variation of velocity in the propagation of the pressure wave generated by the passage of blood flow in vessels. In particular, PWV is strongly connected with variation of vessels diameters and so it can be used to estimate pressure. In this thesis work, a statistical model was implemented to estimate pressure from data collection(ECG, PPG and ABP signals) taken at first from online databases (MIMIC III) and then recorded with State of the art devices (Shimmers and Omron HeartGuide). From electrocardiogram (ECG) and photoplethysmography (PPG) the Pulse Transit Time (PTT) and Heart Rate (HR) are identified in order to train a linear regression algorithm that allows to estimate, after calibration, the systolic and diastolic pressure. The aim of this study is to develop an algorithm capable of accurately predict continuous BP using a Mathematical auto regressive approach that can be integrated in a cuff-less, no-invasive devices more comfortable compared to those currently used.The results obtained show an effective correlation between PTT/HR and systolic and diastolic BP. The main issue of this approach is that it is person-specific and could change with aging. Besides, this problem could be easily overcome with periodic calibration.