Image-Based Artificial Intelligence Methods for Product Control of Tablet Coating Quality

Abstract

Mimicking the human decision-making process is challenging. Especially, many process control situations during the manufacturing of pharmaceuticals are based on visual observations and related experience-based actions. The aim of the present work was to investigate the use of image analysis to classify the quality of coated tablets. Tablets with an increasing amount of coating solution were imaged by fast scanning using a conventional office scanner. A segmentation routine was implemented to the images, allowing the extraction of numeric image-based information from individual tablets. The image preprocessing was performed prior to utilization of four different classification techniques for the individual tablet images. The support vector machine (SVM) technique performed superior compared to a convolutional neural network (CNN) in relation to computational time, and this approach was also slightly better at classifying the tablets correctly. The fastest multivariate method was partial least squares (PLS) regression, but this method was hampered by the inferior classification accuracy of the tablets. Finally, it was possible to create a numerical threshold classification model with an accuracy comparable to the SVM approach, so it is evident that there exist multiple valid options for classifying coated tablets.

Applications of Artificial Intelligence and Machine Learning Algorithms to Crystallization

“…Other physical properties of interest at this stage are the disintegration time and drug release profiles, which can be evaluated from the proportions of excipients, coating level of tablets, or NIR and Raman data with techniques ranging from ANN to SVM and ensemble of regression trees. , In addition, the physical stability of 646 state dispersions of 50 different drugs was classified by eight ML algorithms achieving a prediction accuracy of 0.82 using DNN and RF, potentially saving months in the product development pipeline of water-insoluble drugs . Moreover, SVM was compared to and deemed better than ANN and PLSR to classify tablet coating quality, which uses image-based data to optimize the coating process parameters . On a practical process development level, manufacturing protocols often rely on decisions based on heuristics, which have the potential for enhancement using information from previously collected data.…”

“…270 Moreover, SVM was compared to and deemed better than ANN and PLSR to classify tablet coating quality, which uses image-based data to optimize the coating process parameters. 271 On a practical process development level, manufacturing protocols often rely on decisions based on heuristics, which have the potential for enhancement using information from previously collected data. For instance, Kumar et al 177 used four ML algorithms trained with discrete element simulations to develop a protocol to ensure uniform solid bed mixing of pharmaceutical powders in agitated filter dryers, finding regression tree random forest giving the best accuracy.…”

Image Analysis: A Versatile Tool in the Manufacturing and Quality Control of Pharmaceutical Dosage Forms

“…According to their results, using an appropriate algorithm, the tablets could be classified according to the perfection of the coating [97].…”

“…Hirschberg and colleagues used a simple office scanner for tablet coating quality control. According to their results, using an appropriate algorithm, the tablets could be classified according to the perfection of the coating [ 97 ].…”

Smartphone-enabled 3D printing of medicines

“…An additional benefit of 3D printing is that it is digital; dosage forms are designed in CAD software and digitised prior to printing (Eleftheriadis et al, 2021a). This digital architecture also means it is relatively straightforward to combine 3D printing manufacture with other cybernated technologies, such as artificial intelligence (AI) (Hirschberg et al, 2020), machine learning (Elbadawi et al, 2020a;Elbadawi et al, 2020b), finite element analysis (Karavasili et al, 2020) and mobile applications (Arden et al, 2021;Muñiz Castro et al, 2021), and it is this combination of digital approaches that really offers the potential to reshape the pharmaceutical landscape.…”