Implementation strategy of insulin therapy under sparse state feedback

Traditional continuous glucose monitoring is often costly and inconvenient, necessitating more efficient methods. This paper proposes a novel approach to diabetes management by utilizing sparse monitoring data for insulin injection decisions. A differential equation-based model is developed to estimate plasma glucose levels and optimize insulin dosages and injection intervals using limited feedback data. This method ensures that glucose levels remain within a safe range. Numerical simulations demonstrate the effectiveness of this approach, offering a viable alternative for improving diabetes management.