Leveraging artificial intelligence for cardiovascular risk: a primary care perspective

Cardiovascular diseases (CVDs) remain the leading cause of mortality worldwide. Primary care physicians, particularly family doctors, play a pivotal role in early detection and prevention. This study investigates the potential of artificial intelligence (AI)-based temporal deep learning (DL) models to support cardiovascular risk (CVR) stratification in primary care.

We implemented temporal DL architectures, namely Long Short-Term Memory (LSTM) network and Gated Recurrent Unit (GRU), on a clinically realistic synthetic patient cohort. The dataset included demographic and clinical variables such as age, sex, body mass index (BMI), blood pressure (BP), and established risk factors (smoking, diabetes, and hypertension). Models were trained on sequential data to predict CVR categories (low, moderate, and very high) along with their corresponding probabilities. Model outputs were statistically benchmarked and subsequently aligned with actionable clinical recommendations.

Both LSTM and GRU models demonstrated the ability to forecast CVR across multiple time horizons. Predictions were successfully translated into clinically interpretable recommendations supporting tailored follow-up intervals and targeted interventions.

Integration of AI-driven CVR forecasting into routine family medicine practice can enhance early intervention strategies, optimize patient management, and improve resource allocation. This approach highlights the potential of temporal DL to strengthen preventive care outcomes in primary care settings.