Machine vision-based non-destructive dissolution prediction of meloxicam-containing tablets.

Area of Science:

• Pharmaceutical Technology
• Analytical Chemistry
• Image Analysis

Background:

• Machine vision systems are crucial for pharmaceutical solid dosage form quality assessment.
• These systems support continuous manufacturing, process analytical technology (PAT), quality-by-design (QbD), and real-time release testing (RTRT).

Purpose of the Study:

• To develop a digital UV/VIS imaging system for predicting the in vitro dissolution of meloxicam tablets.
• To non-destructively predict critical process parameters affecting dissolution using multivariate analysis of UV/VIS images.

Main Methods:

• Utilized UV/VIS imaging to capture tablet data.
• Applied multivariate analysis and artificial neural networks (ANNs) for dissolution profile prediction.
• Employed partial least squares (PLS) models based on UV/VIS image color components to predict API content.

Main Results:

• Achieved a prediction error (RMSE) of less than 5% for dissolution profile points.
• Predicted alterations in API content with a relative error of less than 10%.
• Demonstrated that API content directly impacts maximum concentrations observed during dissolution.

Conclusions:

• Presents a novel, non-destructive PAT solution for real-time tablet dissolution testing.
• Highlights the potential of UV/VIS imaging and machine learning for pharmaceutical quality assurance.
• Enables enhanced process understanding and control in tablet manufacturing.