Automated tablet defect detection and the prediction of disintegration time and crushing strength with deep learning based on tablet surface images.

This paper presents novel measurement methods, where deep learning was used to detect tableting defects and determine the crushing strength and disintegration time of tablets on images captured by machine vision. Five different classes of defects were used and the accuracy of the real-time defect recognition performed with the deep learning algorithm YOLOv5 was 99.2%. The system can already match the production capability of tablet presses, with still further room left for improvement. The YOLOv5 algorithm was also used to determine the disintegration time and crushing strength of tablets produced at different compression force settings based on their surface texture. With these accurate, low-cost methods, the 100% screening of the produced tablets could be carried out, resulting in the improvement of quality control and effectiveness of pharmaceutical production.

Process analytical technology based quality assurance of API concentration and fiber diameter of electrospun amorphous solid dispersions.

In this study, a novel quality assurance system was developed utilizing Process analytical technology (PAT) tools and artificial intelligence (AI). Our goal was to monitor the critical quality attributes (CQAs) like drug concentration, morphology and fiber diameter of electrospun amorphous solid dispersion (ASD) formulations with fast at-line techniques. Doxycycline-hyclate (DOX), a tetracycline-type antibiotic was used as a model drug with 2-hydroxypropyl-ß-cyclodextrin (HP-ß-CD) as the matrix excipient. The water-based formulations were electrospun with high-speed electrospinning (HSES). Raman and NIR sensors and machine vision-based color measurement techniques were employed to accurately determine the drug concentration. Given that morphology can influence the solubility of the drug, a convolutional neural network (CNN)-based AI model was developed to examine this property and detect manufacturing defects. Additionally, the diameter of electrospun fibrous samples was measured using camera images and a trained AI model, enabling rapid analysis of fiber diameter with results similar to that of scanning electron microscopy (SEM). These methods and models demonstrate potential in-line analytical tools, offering rapid, cheap and non-destructive analysis of ASD formulations.

Evaluation of floatability characteristics of gastroretentive tablets using VIS imaging with artificial neural networks.

Gastroretentive dosage forms are recommended for several active substances because it is often necessary for the drug to be released from the carrier system into the stomach over an extended period. Among gastroretentive dosage forms, floating tablets are a very popular pharmaceutical technology. In this study, it was investigated whether a rapid, nondestructive method can be used to characterize the floating properties of a tablet. To accomplish our objective, the same composition was compressed, and varied compression forces were applied to achieve the desired tablet. In addition to physical examinations, digital microscopic images of the tablets were captured and analyzed using image analysis techniques, allowing the investigation of the floatability of the dosage form. Image processing algorithms and artificial neural networks (ANNs) were utilized to classify the samples based on their strength and floatability. The input dataset consisted solely of the acquired images. It has been shown by our research that visible imaging coupled with pattern recognition neural networks is an efficient way to categorize these samples based on their floatability. Rapid and non-destructive digital imaging of tablet surfaces is facilitated by this method, offering insights into both crushing strength and floating properties.

Development and integration of a continuous horizontal belt filter into drug production procedure.

In the pharmaceutical industry, filtration is traditionally carried out in batch mode. However, with the spread of continuous technologies, there is an increasing demand for robust continuous filtration strategies suitable for processing suspensions produced in continuous crystallizers. Accordingly, this study aimed to investigate a lab-scale horizontal conveyor belt filtration approach for pharmaceutical separation purposes for the first time. The newly developed continuous horizontal belt filter (CHBF) was tested under different systems (microcrystalline cellulose (MCC)/water, lactose/ethanol and acetylsalicylic acid (ASA)/water) and diverse conditions. Filtration was robust using a well-defined unimodal particle size distribution MCC in water system, where the residual moisture content varied within narrow limits of 45-52% independently from the process conditions. Besides, the residual moisture content highly depended on the applied solvent and particle size. It could be reduced to below 2% by processing the suspensions of either a volatile solvent (lactose in ethanol) or an aqueous slurry of a large particle size ASA. Finally, the CHBF was connected to a mixed suspension mixed product removal (MSMPR) or a plug flow crystallizer (PFC). The residual moisture content of the CHBF-filtered ASA product and operation characteristics (onset of steady-state) were evaluated in both continuous crystallizer-filter systems. The MSMPR-CHBF system operated with a longer startup period. The size of the in situ-produced crystals was of a similar order magnitude in both systems, resulting in a similar residual moisture content (around 20%). Overall, the tested continuous filter was robust, did not modify the crystal morphology in the examined experimental range, and could be effectively integrated with continuous crystallizers.

In-line indirect concentration measurement of ultralow dose API during twin-screw wet granulation based on NIR and Raman spectroscopy.

Twin-screw wet granulation (TWSG) is a promising continuous alternative of pharmaceutical wet granulation. One of its benefits is that the components dissolved in the granulation liquid are distributed homogeneously in the granules. This provides an elegant way to manufacture products with ultralow drug doses. Near-infrared (NIR) and Raman spectroscopy are well-established process analytical technology (PAT) tools that can be used for the in-line monitoring of TSWG. However, their detection limit does not enable the measurement of components in the ultralow (i.e., ppm) range. In this paper, an indirect approach is presented that enables the real-time determination of the concentration of a drug in concentrations between 40 and 100 ppm by using the signal of an excipient, in this case, the polyvinylpyrrolidone (PVP). This component is also dissolved in the granulation liquid; therefore, it is distributed in the same way as the active ingredient. Results of HPLC measurements have proved that the models trained to quantify the concentration of PVP in real-time gave an accurate determination for the active ingredient as well (root mean squared error was 7.07 ppm for Raman and 5.31 ppm for NIR spectroscopy, respectively). These findings imply that it is possible to indirectly predict the concentration of ultralow dose drugs with in-line analytical techniques based on the concentration of an excipient.

UV/VIS-imaging of white caffeine tablets for prediction of CQAs: API content, crushing strength, friability, disintegration time and dissolution profile.

The paper provides a demonstration of how UV/VIS imaging can be employed to evaluate the crushing strength, friability, disintegration time and dissolution profile of tablets comprised of solely white components. The samples were produced using different levels of compression force and API content of anhydrous caffeine. Images were acquired from both sides of the samples using UV illumination for the API content prediction, while the other parameters were assessed using VIS illumination. Based on the color histograms of the UV images, API content was predicted with 5.6 % relative error. Textural analysis of the VIS images yielded crushing strength predictions under 10 % relative error. Regarding friability, three groups were established according to the weight loss of the samples. Likewise, the evaluation of disintegration time led to the identification of three groups: <10 s, 11-35 s, and over 36 s. Successful classification of the samples was achieved with machine learning algorithms. Finally, immediate release dissolution profiles were accurately predicted under 5 % of RMSE with an artificial neural network. The 50 ms exposition time during image acquisition and the resulting outcomes underscore the practicality of machine vision for real-time quality control in solid dosage forms, regardless of the color of the API.

Explainable deep recurrent neural networks for the batch analysis of a pharmaceutical tableting process in the spirit of Pharma 4.0.

Due to the continuously increasing Cost of Goods Sold, the pharmaceutical industry has faced several challenges, and the Right First-Time principle with data-driven decision-making has become more pressing to sustain competitiveness. Thus, in this work, three different types of artificial neural network (ANN) models were developed, compared, and interpreted by analyzing an open-access dataset from a real pharmaceutical tableting production process. First, the multilayer perceptron (MLP) model was used to describe the total waste based on 20 raw material properties and 25 statistical descriptors of the time series data collected throughout the tableting (e.g., tableting speed and compression force). Then using 10 process time series data in addition to the raw material properties, the cumulative waste, during manufacturing was also predicted by long short-term memory (LSTM) and bidirectional LSTM (biLSTM) recurrent neural networks (RNN). The LSTM network was used to forecast the waste production profile to allow preventive actions. The results showed that RNNs were able to predict the waste trajectory, the best model resulting in 1096 and 2174 tablets training and testing root mean squared errors, respectively. For a better understanding of the process, and the models and to help the decision-support systems and control strategies, interpretation methods were implemented for all ANNs, which increased the process understanding by identifying the most influential material attributes and process parameters. The presented methodology is applicable to various critical quality attributes in several fields of pharmaceutics and therefore is a useful tool for realizing the Pharma 4.0 concept.

Real-time release testing of in vitro dissolution and blend uniformity in a continuous powder blending process by NIR spectroscopy and machine vision.

Continuous manufacturing is gaining increasing interest in the pharmaceutical industry, also requiring real-time and non-destructive quality monitoring. Multiple studies have already addressed the possibility of surrogate in vitro dissolution testing, but the utilization has rarely been demonstrated in real-time. Therefore, in this work, the in-line applicability of an artificial intelligence-based dissolution surrogate model is developed the first time. NIR spectroscopy-based partial least squares regression and artificial neural networks were developed and tested in-line and at-line to assess the blend uniformity and dissolution of encapsulated acetylsalicylic acid (ASA) - microcrystalline cellulose (MCC) powder blends in a continuous blending process. The studied blend is related to a previously published end-to-end manufacturing line, where the varying size of the ASA crystals obtained from a continuous crystallization significantly affected the dissolution of the final product. The in-line monitoring was suitable for detecting the variations in the ASA content and dissolution caused by the feeding of ASA with different particle sizes, and the at-line predictions agreed well with the measured validation dissolution curves (f2 = 80.5). The results were further validated using machine vision-based particle size analysis. Consequently, this work could contribute to the advancement of RTRT in continuous end-to-end processes.

UV imaging for the rapid at-line content determination of different colourless APIs in their tablets with artificial neural networks.

This paper presents a novel high-resolution and rapid (50 ms) UV imaging system, which was used for at-line, non-destructive API content determination of tablets. For the experiments, amlodipine and valsartan were selected as two colourless APIs with different UV induced fluorescent properties according to the measured solid fluorescent spectra. Images were captured with a LED-based UV illumination (385-395 nm) of tablets containing amlodipine or valsartan and common tableting excipients. Blue or green colour components from the RGB colour space were extracted from the images and used as an input dataset to execute API content prediction with artificial neural networks. The traditional destructive, solution-based transmission UV measurement was applied as reference method. After the optimization of the number of hidden layer neurons it was found that the relative error of the content prediction was 4.41 % and 3.98 % in the case of amlodipine and valsartan containing tablets respectively. The results open the possibility to use the proposed UV imaging-based system as a rapid, in-line tool for 100 % API content screening in order to greatly improve pharmaceutical quality control and process understanding.

Raman and NIR spectroscopy-based real-time monitoring of the membrane filtration process of a recombinant protein for the diagnosis of SARS-CoV-2.

This research shows the detailed comparison of Raman and near-infrared (NIR) spectroscopy as Process Analytical Technology tools for the real-time monitoring of a protein purification process. A comprehensive investigation of the application and model development of Raman and NIR spectroscopy was carried out for the real-time monitoring of a process-related impurity, imidazole, during the tangential flow filtration of Receptor-Binding Domain (RBD) of the SARS-CoV-2 Spike protein. The fast development of Raman and NIR spectroscopy-based calibration models was achieved using offline calibration data, resulting in low calibration and cross-validation errors. Raman model had an RMSEC of 1.53 mM, and an RMSECV of 1.78 mM, and the NIR model had an RMSEC of 1.87 mM and an RMSECV of 2.97 mM. Furthermore, Raman models had good robustness when applied in an inline measurement system, but on the contrary NIR spectroscopy was sensitive to the changes in the measurement environment. By utilizing the developed models, inline Raman and NIR spectroscopy were successfully applied for the real-time monitoring of a process-related impurity during the membrane filtration of a recombinant protein. The results enhance the importance of implementing real-time monitoring approaches for the broader field of diagnostic and therapeutic protein purification and underscore its potential to revolutionize the rapid development of biological products.

Effects of omitting titanium dioxide from the film coating of a pharmaceutical tablet - An industrial case study of attempting to comply with EU regulation 2022/63.

Recently, concerns have been raised about the safety of titanium dioxide (TiO2), a commonly used component of pharmaceutical film coatings. The European Union has recently prohibited the application of this material in the food industry, and it is anticipated that the same will happen in the pharmaceutical industry. For this reason, pharmaceutical manufacturers have to consider the possible impact of removing TiO2 from the film coating of tablets. In this paper, we present a case study of a commercially produced tablet where the film coating containing TiO2 was replaced with a coating using calcium carbonate (CaCO3) or with a transparent coating. The performance of the coatings was compared by measuring the moisture absorption rate and the dissolution profile of the tablets. In these regards, there were negligible differences between the coating types. The tablets contained a highly photosensitive drug, the ability of the coatings to protect the drug was evaluated through environmental stability and photostability measurements. The HPLC results showed that the inclusion of TiO2 does not provide additional benefits, when humidity and thermal stress is applied, however its role was vital in protecting the drug from external light. There were several decomposition products which appeared in large quantities when TiO2 was missing from the coating. These results imply that photosensitivity is an issue, replacing TiO2 will be challenging, though its absence can be tolerated when the drug does not need to be protected from light.

UV-VIS imaging-based investigation of API concentration fluctuation caused by the sticking behaviour of pharmaceutical powder blends.

Surface powder sticking in pharmaceutical mixing vessels poses a risk to the uniformity and quality of drug formulations. This study explores methods for evaluating the amount of pharmaceutical powder mixtures adhering to the metallic surfaces. Binary powder blends consisting of amlodipine and microcrystalline cellulose (MCC) were used to investigate the effect of the mixing order on the adherence to the vessel wall. Elevated API concentrations were measured on the wall and within the dislodged material from the surface, regardless of the mixing order of the components. UV imaging was used to determine the particle size and the distribution of the API on the metallic surface. The results were compared to chemical maps obtained by Raman chemical imaging. The combination of UV and VIS imaging enabled the rapid acquisition of chemical maps, covering a substantially large area representative of the analysed sample. UV imaging was also applied in tablet inspection to detect tablets that fail to meet the content uniformity criteria. The results present powder adherence as a possible source of poor content uniformity, highlighting the need for 100% inspection of pharmaceutical products to ensure product quality and safety.

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

Machine vision systems have emerged for quality assessment of solid dosage forms in the pharmaceutical industry. These can offer a versatile tool for continuous manufacturing while supporting the framework of process analytical technology, quality-by-design, and real-time release testing. The aim of this work is to develop a digital UV/VIS imaging-based system for predicting the in vitro dissolution of meloxicam-containing tablets. The alteration of the dissolution profiles of the samples required different levels of the critical process parameters, including compression force, particle size and content of the API. These process parameters were predicted non-destructively by multivariate analysis of UV/VIS images taken from the tablets. The dissolution profile prediction was also executed using solely the image data and applying artificial neural networks. The prediction error (RMSE) of the dissolution profile points was less than 5%. The alteration of the API content directly affected the maximum concentrations observed at the end of the dissolution tests. This parameter was predicted with a relative error of less than 10% by PLS models that are based on the color components of UV and VIS images. In conclusion, this paper presents a modern, non-destructive PAT solution for real-time testing of the dissolution of tablets.

Quantitative and qualitative analysis of cell culture media powders for mammalian cells by Raman microscopy.

Cell culture media are essential for large-scale recombinant protein production using mammalian cell cultures. The composition and quality of media significantly impact cell growth and product formation. Analyzing media poses challenges due to complex compositions and undisclosed exact compositions. Traditional methods like NMR and chromatography offer sensitivity but require time-consuming sample preparation and lack spatial information. Raman chemical mapping characterizes solids, but its use in cell culture media analysis is limited so far. We present a chemometric evaluation for Raman maps to qualify and quantify media components, evaluate powder homogeneity, and perform lot-to-lot comparisons. Three lots of a marketed cell culture media powder were measured with Raman mapping technique. Chemometrics techniques have outlined a strategy to extract information from complex data. First, a spectral library has been structured. In addition to the 23 spectra for presumed ingredients, we obtained another 9 pure components with Multivariate Curve Resolution-Alternating Least Squares (MCR-ALS). Then the Spectral Angle Mapper-Orthogonal Projection (SAM-OP) algorithm revealed whether references actually occur in the mapped media powders. Finally, a quantification was provided by Classical Least Squares (CLS) modelling. Quantities of 18 significant amino acids mostly correlated with the reference method. The proposed method can be generally applied even for such complicated samples. Leveraging Raman mapping and innovative chemometric methods enhance recombinant protein production by improving the understanding of the spatial distribution and composition of cell culture media in mammalian cell cultivations.

Development of a PAT platform for the prediction of granule tableting properties.

In this work, the feasibility of implementing a process analytical technology (PAT) platform consisting of Near Infrared Spectroscopy (NIR) and particle size distribution (PSD) analysis was evaluated for the prediction of granule downstream processability. A Design of Experiments-based calibration set was prepared using a fluid bed melt granulation process by varying the binder content, granulation time, and granulation temperature. The granule samples were characterized using PAT tools and a compaction simulator in the 100-500 kg load range. Comparing the systematic variability in NIR and PSD data, their complementarity was demonstrated by identifying joint and unique sources of variation. These particularities of the data explained some differences in the performance of individual models. Regarding the fusion of data sources, the input data structure for partial least squares (PLS) based models did not significantly impact the predictive performance, as the root mean squared error of prediction (RMSEP) values were similar. Comparing PLS and artificial neural network (ANN) models, it was observed that the ANNs systematically provided superior model performance. For example, the best tensile strength, ejection stress, and detachment stress prediction with ANN resulted in an RMSEP of 0.119, 0.256, and 0.293 as opposed to the 0.180, 0.395, and 0.430 RMSEPs of the PLS models, respectively. Finally, the robustness of the developed models was assessed.

Development and Comparison of Alternative Methods for the Purification of Adalimumab Directly from Harvested Cell Culture Fluid.

Research background: Protein A affinity chromatography is a well-established method currently used in the pharmaceutical industry. However, the high costs usually associated with chromatographic separation of protein A and the difficulties in continuous operation make the investigation of alternative purification methods very important. Experimental approach: In this study, extraction/back-extraction and precipitation/dissolution methods were developed and optimised. They were compared with protein A and cation exchange chromatography separations in terms of yield of monoclonal antibody (mAb) and amount of residual impurities, such as DNA and host cell proteins, and amount of mAb aggregates. For a comprehensive comparison of the different methods, experiments were carried out with the same cell-free fermentation broth containing adalimumab. Results and conclusions: Protein A and cation exchange chromatographic separations resulted in high yield and purity of adalimumab. The precipitation-based process resulted in high yield but with lower purity. The extraction-based purification resulted in low yield and purity. Thus, the precipitation-based method proved to be more promising than the extraction-based method for direct purification of adalimumab from harvested cell culture fluid. Novelty and scientific contribution: Although alternative purification methods may offer the advantages of simplicity and low-cost operation, further significant improvements are required to compete with the performance of chromatographic separations of adalimumab from true fermentation broth.

Image-based simultaneous particle size distribution and concentration measurement of powder blend components with deep learning and machine vision.

This work presents a system, where deep learning was used on images captured with a digital camera to simultaneously determine the API concentration and the particle size distribution (PSD) of two components of a powder blend. The blend consisted of acetylsalicylic acid (ASA) and calcium hydrogen phosphate (CHP), and the predicted API concentration was found corresponding with the HPLC measurements. The PSDs determined with the method corresponded with those measured with laser diffraction particle size analysis. This novel method provides fast and simple measurements and could be suitable for detecting segregation in the powder. By examining the powders discharged from a batch blender, the API concentrations at the top and bottom of the container could be measured, yielding information about the adequacy of the blending and improving the quality control of the manufacturing process.

Comparing the Performance of Raman and Near-Infrared Imaging in the Prediction of the In Vitro Dissolution Profile of Extended-Release Tablets Based on Artificial Neural Networks.

In this work, the performance of two fast chemical imaging techniques, Raman and near-infrared (NIR) imaging is compared by utilizing these methods to predict the rate of drug release from sustained-release tablets. Sustained release is provided by adding hydroxypropyl methylcellulose (HPMC), as its concentration and particle size determine the dissolution rate of the drug. The chemical images were processed using classical least squares; afterwards, a convolutional neural network was applied to extract information regarding the particle size of HPMC. The chemical images were reduced to an average HPMC concentration and a predicted particle size value; these were used as inputs in an artificial neural network with a single hidden layer to predict the dissolution profile of the tablets. Both NIR and Raman imaging yielded accurate predictions. As the instrumentation of NIR imaging allows faster measurements than Raman imaging, this technique is a better candidate for implementing a real-time technique. The introduction of chemical imaging in the routine quality control of pharmaceutical products would profoundly change quality assurance in the pharmaceutical industry.

In-line particle size measurement during granule fluidization using convolutional neural network-aided process imaging.

This paper presents a machine learning-based image analysis method to monitor the particle size distribution of fluidized granules. The key components of the direct imaging system are a rigid fiber-optic endoscope, a light source and a high-speed camera, which allow for real-time monitoring of the granules. The system was implemented into a custom-made 3D-printed device that could reproduce the particle movement characteristic in a fluidized-bed granulator. The suitability of the method was evaluated by determining the particle size distribution (PSD) of various granule mixtures within the 100-2000 µm size range. The convolutional neural network-based software was able to successfully detect the granules that were in focus despite the dense flow of the particles. The volumetric PSDs were compared with off-line reference measurements obtained by dynamic image analysis and laser diffraction. Similar trends were observed across the PSDs acquired with all three methods. The results of this study demonstrate the feasibility of performing real-time particle size analysis using machine vision as an in-line process analytical technology (PAT) tool.

Integrated continuous melt granulation-based powder-to-tablet line: Process investigation and scale-up on the same equipment.

In the last decades, continuous manufacturing (CM) has become a research priority in the pharmaceutical industry. However, significantly fewer scientific researches address the investigation of integrated, continuous systems, a field that needs further exploration to facilitate the implementation of CM lines. This research outlines the development and optimization of an integrated, polyethylene glycol aided melt granulation-based powder-to-tablet line that operates fully continuously. The flowability and tabletability of a caffeine-containing powder mixture were improved through twin-screw melt granulation resulting in the production of tablets with improved breaking force (from 15 N to over 80 N), excellent friability, and immediate release dissolution. The system was also conveniently scaleable: the production speed could be increased from 0.5 kg/h to 8 kg/h with only minimal changes in the process parameters and using the same equipment. Thereby the frequent challenges of scale-up can be avoided, such as the need for new equipment and separate optimization.