Optimal experiment design with applications to Pharmacokinetic modeling

Area of Science:

• Dynamical Systems and Control
• Pharmacokinetics and Pharmacodynamics
• Mathematical Modeling and Optimization

Background:

• Accurate estimation of physiological parameters in pharmacokinetic models is crucial for drug development and personalized medicine.
• Traditional input design methods may not be optimal for limited measurement scenarios.
• Safety constraints, such as injection rates and dosage, must be considered in experimental design.

Purpose of the Study:

• To develop a method for designing optimal inputs for dynamical systems to enhance parameter estimation.
• To apply this method to pharmacokinetic model parameter identification using limited data.
• To optimize the input design considering safety and experimental constraints.

Main Methods:

• Utilizing A and D optimality criteria based on the Fisher Information Matrix for optimization.
• Implementing a two-stage approach: an initial learning phase and a subsequent optimization phase.
• Defining model inputs as intravenous drug injections with constraints on injection rates and total dosage.

Main Results:

• A strategy for designing optimal experimental inputs was formulated.
• The proposed method allows for improved estimation of pharmacokinetic parameters with limited measurements.
• The two-stage approach effectively balances initial model learning with targeted parameter optimization.

Conclusions:

• The developed input design methodology enhances parameter estimation accuracy in dynamical systems, particularly for pharmacokinetic applications.
• The staged approach, incorporating learning and optimization, is effective under safety and data constraints.
• This work provides a framework for optimizing experimental design in complex biological systems.