A Telemonitoring Solution Applied to Dynamic CVD Risk Assessment: LookAfterRisk

Cardiovascular disease (CVD) risk assessment models usually do not incorporate any dynamic information. In fact, the risk of a specific CVD event (e.g., death and rehospitalization) of a patient is assessed exclusively based on the values of a set of features at a given moment not considering the impact of their evolution.The LookAfterRisk project intends to overcome this aspect, through the integration into the CVD risk assessment of data collected at the patient's home during a certain period of time. Thus, the developed models are applied at hospital admission, in patients with a first episode of acute myocardial infarction (MI) and are continuously updated during the follow-up period when the patient returns home. A technical platform comprising low cost home-mobile technologies must also be developed to obtain the required patient's data as well as to provide the respective risk level to the physicians. Therefore, the main goal is the development of dynamic models for CVD risk assessment of acute events integrating data from remote monitoring, in order to stratify patients according to their care needs (management of MI patients). In parallel, recent advances in medicine and in technology infrastructures allow the collection of important information. As result, three main scientific challenges will be addressed throughout this project: i) knowledge discovery/extraction based on clinical datasets; ii) fusion of that knowledge with current clinical knowledge; iii) dynamic update of risk level based on data collected at patient home during the follow up period.From the technical perspective, a platform should be implemented in order to accomplish three main requirements: i) integration of the developed data analysis algorithms; ii) access to the patient's data stored in the Hospital Information System as well as to assure the interaction with the physician (provide the risk level/risk evolution); iii) integration of the mobile application to monitor the blood pressure and heart rate. The validation of initial models comprises two phases: i) based on the Portuguese Society of Cardiology (PSC) dataset of MI patients (N=16000); ii) based on a home telemonitoring study with a follow-up of 9 months, including 50 patients (selected by the clinical partner from patients with a first episode of acute MI at the Leiria Hospital).