Scientific papers
Understanding the mechanism behind the disintegration of pharmaceutical tablets remains largely incomplete. Despite the crucial need to regulate the disintegration process to optimize the bioavailability of active pharmaceutical ingredients and ensure consistent release patterns, our current comprehension relies on indirect or surface-level assessments. Consequently, formulation design often relies heavily on empirical knowledge, leading to unpredictability and inefficiency.
Our objective is to narrow this knowledge gap by utilizing time-resolved X-ray micro-computed tomography (µCT) to capture volumetric images of various mini-tablet formulations undergoing disintegration. Overcoming the challenge of analyzing heterogeneous tomographic images at high magnification necessitated automated image segmentation. We developed and trained a convolutional neural network (CNN) based on the U-Net architecture for rapid and consistent image segmentation. Additionally, we established a µCT data reconstruction pipeline tailored to optimize image quality for CNN-based segmentation.
Through our methodology, we gained insights into the internal microstructures of disintegrating tablets and extracted parameters related to disintegration kinetics from the time-resolved data. Utilizing factor analysis, we identified the impact of different formulation components on the disintegration process, evaluating qualitative and quantitative experimental responses. Our findings were contextualized with known properties of formulation components and established experimental outcomes. Leveraging deep learning-based image processing, our direct imaging approach provided novel perspectives into the disintegration mechanism of pharmaceutical tablets.
Comments
No comments posted yet.
Add a comment
Articles you may like
Related products
Looking for additional information?
Let us help you