Descrizione del progetto
Un’intelligenza artificiale all’avanguardia per prevenire le malattie cardiache e il diabete
Le malattie non trasmissibili come le cardiopatie e il diabete sono in aumento e rappresentano una grande sfida per la sanità pubblica. Il progetto SHIELD, finanziato dall’UE, si propone di affrontare il problema usando l’intelligenza artificiale per prevenire queste patologie. SHIELD raccoglie dati multimodali da diversi attori sanitari per creare punteggi di rischio e strategie di prevenzione personalizzate. Questi interventi, sperimentati in Spagna, Italia e Svizzera, vengono erogati tramite un’applicazione mobile e comprendono obiettivi di cambiamento del comportamento, consigli sullo stile di vita e contenuti di alfabetizzazione sanitaria per ridurre il rischio e la progressione delle malattie. Basando i piani di cura su metriche di rischio individuali, SHIELD vuole migliorare la prevenzione e la gestione delle malattie cardiovascolari e del diabete.
Obiettivo
The prevalence of Non-Communicable Diseases is forcing countries to consider different initiatives aimed at reducing the burden and impact of these diseases.
In this sense, SHIELD pioneers an innovative approach to preventing cardiovascular diseases (CVD) and diabetes at all stages of disease and considering the strong relationship among them. Utilizing advanced AI, SHIELD offers personalized interventions following a hierarchical model based on patients' risk profiles. These profiles are continuously assessed through risk stratification and disease progression tools, leading to low, moderate, and high-risk layers, each requiring distinct prevention strategies, from halting disease progression to preventing relapse and complications.
Initial risk assessment will include genetics, demographics, socio-economic status, environment, behavior, and medical conditions, using datasets like SHARE or ELSA, and retrospective hospital data (4,500+ patients). Later-stage disease analysis will also entail polypharmacy, treatment adherence, wearables, psychosocial factors, PROMs, and PREMs, allowing for individualized interventions and real-time alerts for professionals through the SHIELD dashboard. SHIELD also prioritizes the quality and security of these data, creating a standardized data homogenization model and federated learning approach to keep sensitive data locally. Transparency is provided through explainable ML tools.
Interventions are accessible via mobile apps, providing resources, recommendations, education, and local services. Optimization algorithms enhance user engagement, and a chatbot, trained on patient data, offers continuous support with professional oversight.
SHIELD will be validated in 3 pilots, involving more than 2,300 individuals along 2 years. The aim is to get knowledge on the cost-effectiveness and efficacy of the prevention strategy proposed by SHIELD and to get insights for effective primary prevention pathways with population-wide impact.
Parole chiave
Programma(i)
Invito a presentare proposte
(si apre in una nuova finestra) HORIZON-HLTH-2024-STAYHLTH-01-two-stage
Vedi altri progetti per questo bandoMeccanismo di finanziamento
HORIZON-RIA -Coordinatore
28036 Madrid
Spagna
L’organizzazione si è definita una PMI (piccola e media impresa) al momento della firma dell’accordo di sovvenzione.