Patient-Specific Modeling and Model Predictive Control Approach to Personalized Optimal Anemia Management

Area of Science:

• Biomedical Engineering
• Nephrology
• Control Systems

Background:

• Anemia is common in chronic kidney disease (CKD) due to reduced erythropoietin (EPO) production, leading to low hemoglobin (Hgb).
• External EPO administration aims to maintain Hgb levels between 10-12 g/dL, but current methods like the anemia management protocol (AMP) struggle with dosage stability and Hgb oscillations.

Purpose of the Study:

• To develop and evaluate a personalized model predictive control (MPC) framework for optimizing EPO dosage in CKD patients.
• To compare the efficacy of MPC against the existing anemia management protocol (AMP) in stabilizing Hgb levels.

Main Methods:

• A semi-blind robust model identification method was used to create personalized patient models from minimal dose-response data.
• These personalized models were integrated into a model predictive control (MPC) framework for EPO dose guidance.
• Simulations were conducted to compare MPC performance against the anemia management protocol (AMP).

Main Results:

• Model predictive control (MPC) demonstrated superior performance in maintaining target hemoglobin (Hgb) levels (10-12 g/dL) without significant overshoots or undershoots.
• MPC effectively stabilized EPO dosage, unlike the anemia management protocol (AMP), which exhibited oscillations in Hgb levels.
• In-silico results indicate MPC's capability for more precise and stable anemia management in CKD.

Conclusions:

• Model predictive control (MPC), utilizing personalized patient models, offers a significant advancement in managing anemia for chronic kidney disease (CKD) patients.
• This research provides a clinical decision-making framework for personalized EPO dosing, improving upon current anemia management protocols.
• The proposed method uses minimal patient data, making it practical for clinical application.