On the Design of a Smart Mirror for Cardiovascular Risk Prediction

According to the World Health Organization, cardiovascular diseases are one of the first cause of death. Prevention, together with an healthy life style, could reduce these high numbers. Moreover, telehealth systems that allow remote and continuous monitoring of patients can be used for alerting the medical experts if anomalies are detected. In this work, a smart mirror for vital parameters measurements and cardiovascular risk assessment is proposed. It is a contactless solution, based on remote photopletysmography (rPPG) that allows natural and easy measurements. Moreover, a Hierarchical Fuzzy Inference System (HFIS) has been used to predict the cardiovascular risk level. Finally, the acceptability of the new technology has been measured, since in the medical domain users need to trust the results obtained by the automatic techniques. Results have shown that the measurements performed with the smart mirror are reliable, since they are comparable with the pulse oximeter, used as baseline. Moreover, the use of HFIS has significantly reduced the number of rules, while preserving a good accuracy. Finally, Technology Acceptance Model (TAM) has shown high usability and acceptability values, thus suggesting that the smart solution can be adopted in daily routine.