Development of Automated Testing Algorithms for Graphical User Interface and Pulse Wave Velocity Evaluation on LDV Signals

According to the World Health Organization, Cardiovascular diseases are the leading cause of death in the world nowadays. Over the years, numerous approaches have been tried in the attempt to detect and prevent them, resulting in the selection of physiological risk factors responsible for generating these pathologies. Among the various detectors, aortic stiffness is considered one of the most accurate indicators for preventing the development of diseases related to the cardiovascular system. It is currently assessed by measuring Pulse Wave Velocity (PWV) in the carotid-femoral segment and, over the years, efforts have been made to calculate this variable as simply and, above all, non-invasively as possible. As part of the EU-funded “InSiDe” project, a device has been developed which is able to calculate PWV from LDV signals obtained from Laser Doppler technology. The device is equipped with a handset, used by the healthcare provider to detect the LDV signals at two critical points of the patient, and a personal computer with a Graphical User Interface (GUI), where the clinical data of the patient can be entered, and the signals can be visualised during the acquisition. The aim of the present work is the implementation of automatic controls to be integrated within the GUI, as tests, relating to the input of patient data by an operator. With the help of the Unittest library, it was possible to configure the individual tests independently, collectively incorporated inside a single algorithm. The implementation of these controls is crucial in the development of the patient report in order to avoid any unforeseen events or misunderstandings in future consultations. An automatic algorithm was therefore created that would generate confirmation or warning alerts based on the input of each individual physiological parameter. The data and ranges of values used for some parameters were taken from a report compiled by the “InSiDe” project containing the list of checks to be performed in the GUI, for others they were chosen arbitrarily in a reasonable manner, while waiting for feedback and confirmation from the clinical staff. A further aim of the work was to implement a function in the Python environment for evaluating PWV from LDV signals, which would be integrated into the GUI. The displacement signals utilised as dataset came from acquisitions performed with the help of the CARDIS device, a European project preceding the current “InSiDe”, on a total of 100 patients at the Hopital Europen Georges Pompidou (HEGP) in Paris. The technique used in the function to assess PWV is a median of the median PWVs, calculated from the estimated Pulse Transit Time and fiducial points inside the signals. The existing function was designed in the Matlab programming language, the contribution of this thesis work was the implementation of it in the Python environment, so that it could be integrated within the GUI algorithm and allow a more accurate estimation of the signals acquired by the “InSiDe” device.