Application of multi-objective optimization algorithm to the preparation of polycaprolactone microsphere formulations

Area of Science:

• Materials Science
• Biomedical Engineering
• Formulation Science

Background:

• Polycaprolactone microspheres (PCL-MS) are vital for tissue filling applications.
• Particle size and uniformity critically influence PCL-MS efficacy in filling and therapeutic outcomes.
• Efficient optimization strategies are crucial for advancing PCL-MS development.

Purpose of the Study:

• To optimize polycaprolactone microsphere (PCL-MS) preparation using advanced design and intelligent algorithms.
• To develop mathematical models predicting PCL-MS particle size and distribution width.
• To identify optimal PCL-MS formulations for improved therapeutic applications.

Main Methods:

• Box-Behnken design was employed to study PCL concentration, polyvinyl alcohol concentration, and water-oil ratio.
• Mathematical models were developed to predict particle size (Y1) and particle size distribution width (Y2).
• Multi-objective optimization was performed using Nondominated Sorting Genetic Algorithm-II (NSGA-II) and Multi-Objective Artificial Hummingbird Algorithm (MOAHA).

Main Results:

• Two optimal preparation schemes were identified from Pareto solution sets generated by NSGA-II and MOAHA.
• Experimental validation confirmed predicted values for particle size and distribution width with deviations under 5%.
• All optimized protocols met target requirements, demonstrating suitability for PCL-MS preparation.

Conclusions:

• Box-Behnken design combined with intelligent optimization yielded three effective PCL-MS formulations.
• The developed formulations facilitate the production of PCL-MS with reduced particle size and narrower distributions.
• This study significantly advances PCL-MS formulation development for enhanced therapeutic applications.