PROVIDE
Prediction and prevention of cardiovascular diseases in pre- and type 2 diabetes
|
The primary aim of the project is to develop a system for the rapid screening of early CVD risk to improve the quality of clinical monitoring in patients with diabetes, allowing early personalized prevention.
The system is based on a combination of (i) fast, low- cost and widely available technique (electrocardiography); (ii) innovative telehealth technology of portable devices allowing for easy ECG recording tested in a vast amount of data collected by a network of collaborating hospitals in several EU countries; (iii) advanced ECG analysis based on a battery of nonlinear dynamics measures and machine learning models; (iv) measurement of standard markers of diabetes (glucose, HbA1c, lipid, blood pressure, body mass index, waist circumference) and diabetic kidney disease markers (albuminuria, serum creatinine); (iv) lipid metabolomic analyses; (v) novel biomarkers such as adipokines and inflammatory markers. This system will be applied to enable a new understanding of the mechanisms, and to predict individual cardiovascular risk in diabetes. The secondary aim is to improve the classification of patients in sub-types of diabetes at different levels of disease severity and in differentiation between comorbidities that would lead to better clinical decisions and potential remote personalized patient monitoring. The explorative aim of the project is to develop algorithms/decision trees to define and validate diagnostic pathways that tailor therapy towards individual patients’ needs overcoming the “one-size-fits-all” approach. The originality lies both in combination of analytical methods (coming from complex systems dynamics theory, information theory and physicochemical biomarkers), advanced statistical approaches (blended machine learning and Bayesian approach) and application of portable monitoring devices that can lead us to future remote monitoring of ambulatory patients. |
