Using nonlinear model predictive control to find optimal therapeutic strategies to modulate inflammation

Area of Science:

• Critical care medicine
• Computational biology
• Systems immunology

Background:

• Modulating the inflammatory response is crucial for treating critically ill patients.
• Computational models aim to understand systemic inflammation dynamics.
• Translating computational findings into clinical practice requires identifying specific therapeutic targets.

Purpose of the Study:

• To implement a nonlinear model predictive control (NMPC) algorithm for personalized treatment strategies.
• To utilize a reduced computational model of acute inflammatory response to severe infection.
• To identify patient-specific immuno-modulatory approaches for intensive care.

Main Methods:

• Developed and applied a nonlinear model predictive control (NMPC) algorithm.
• Used a reduced computational model simulating acute inflammatory response.
• Incorporated simulated observations of clinically accessible inflammatory mediators.

Main Results:

• NMPC successfully identified patient-specific therapeutic strategies.
• Simulated NMPC strategies outperformed standardized therapies.
• NMPC-derived strategies were superior even to those from general optimization routines.

Conclusions:

• Computational modeling combined with NMPC can suggest novel immuno-modulatory strategies.
• This approach holds practical utility for treating intensive care patients.
• Personalized, model-based interventions show potential for improved clinical outcomes.