Optimization of print parameters for batch and continuous manufacturing of three-dimensional (3D) printed dosage...

Area of Science:

• Pharmaceutical Technology
• Materials Science
• Computational Science

Background:

• Three-dimensional printing (3D printing) offers personalized pharmaceutical oral dosage forms (printlets) with enhanced features like versatile drug release and improved patient compliance.
• Optimizing 3D printing parameters is critical for ensuring the quality of these printlets.
• Machine learning (ML) can significantly reduce development time and costs by optimizing these parameters.

Purpose of the Study:

• To optimize fused deposition modeling (FDM) printing parameters for both batch and continuous manufacturing of pharmaceutical printlets.
• To integrate ML algorithms for predicting and selecting optimal processing parameters to achieve defect-free printlets.
• To validate the ML model's effectiveness across various materials and printing conditions.

Main Methods:

• A three-level full factorial design was used to generate data for training ML algorithms.
• Image segmentation was employed to analyze printlets for defects.
• Gaussian Process Regressor (GPR) and Efficient Global Optimization (EGO) were utilized for parameter prediction and selection.
• The developed algorithm was tested and validated by printing and characterizing printlets.

Main Results:

• The ML algorithm successfully predicted parameter sets for both batch (R²=0.8783) and continuous (R²=0.9364) printing, achieving zero-defect printlets.
• The algorithm demonstrated adaptability to various materials within a 190-220 ℃ temperature range.
• Flow rate was identified as a significant factor influencing printlet quality, more so than print speed, temperature, or infill density.

Conclusions:

• Adaptive ML, specifically GPR and EGO, effectively optimizes FDM printing parameters for high-quality pharmaceutical printlets.
• This ML-driven approach enhances the efficiency and reliability of 3D printing for pharmaceutical dosage manufacturing.
• The study validates the potential of ML in creating defect-free, personalized medicines via 3D printing.