Modeling, design, and machine learning-based framework for optimal injectability of microparticle-based drug...

Area of Science:

• Biopharmaceutical Engineering
• Drug Delivery Systems
• Computational Science

Background:

• Conventional hypodermic needles present significant challenges for the clinical translation of microparticle-based drug formulations.
• Inefficient injection can hinder the effective delivery of biopharmaceuticals and microparticle therapeutics.

Purpose of the Study:

• To identify key factors influencing microparticle injectability.
• To establish a predictive framework for microparticle injection using computational fluid dynamics, design of experiments, and machine learning.
• To design an optimized syringe system for improved microparticle delivery.

Main Methods:

• Development of a numerical multiphysics model to simulate microparticle flow and needle blockage.
• Creation of an empirical mathematical model using experimental data.
• Integration of experimental results into an artificial neural network for predictive modeling.
• Design and testing of a custom injection system.

Main Results:

• A predictive framework was established to assess microparticle injectability.
• The developed empirical model and artificial neural network accurately predicted injection performance.
• A custom injection system demonstrated a sixfold increase in the injectability of large microparticles compared to commercial syringes.
• Optimized syringe design improved in vitro and in vivo injectability.

Conclusions:

• The proposed framework is crucial for optimizing the parenteral delivery of microparticle-based drugs.
• Computational modeling and machine learning offer powerful tools for enhancing drug delivery systems.
• The developed injection system represents a significant advancement for biopharmaceutical drug delivery.