e-Health is a new concept of health management that produces several benefits. First, it reduces sanitary costs by prevention of potential diseases. Besides, it empowers the patients with a new generation of non-invasive, wearable personalized devices to make them more independent, and to provide early signals of health decline and advice for appropriate actions in daily life. Finally, the analysis of the obtained data greatly improves prevention by detecting early patterns of potential diseases; it allows us to evaluate the efficacy of treatments, to understand (through complex processing) the evolution of diseases and the factors that influence them.Biomedical engineers envision “a new system of distributed computing tools that will collect authorized medical data about people and store it securely within a network designed to help deliver quick and efficient care”.
In order to obtain such benefits, the target population has to be monitorized 24 h a day, and a Wireless Body Sensor Network (WBSN) is deployed. Thus, the system is composed of a large set of nodes, distributed among the population. Such nodes are non-intrusive and portable, which impose constraints on their energy consumption. Data obtained by the sensors are communicated to the embedded processing elements (PDAs, smartphones, etc.) by means of wireless connections.
Then, the huge set of data must be analyzed with the aim of performing the epidemiologic assessment. Also, diagnosis algorithms have to be implemented to allow early detection of pathologies and to learn the evolution of patients.
Our research in this area targets all the levels of abstraction of the aforementioned model. Our team works on designing and deploying ultra-low-power WBSNs for ambulatory data acquisition within running clinical trials (in collaboration with national hospitals). The acquired data are processed in a Big Data environment to provide predictive algorithms that anticipate the symptomatic crisis of the patients (such our works on neurological diseases), perform diagnosis classification, allow the simulation of patients, and balance the workload execution among the multiple nodes in a Mobile Cloud Computing scenario.