Computational intelligence modelling of pharmaceutical tabletting processes using bio-inspired optimization algorithms

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.

Deep Fuzzy Min-Max Neural Network: Analysis and Design

Fuzzy min-max neural network (FMNN) is one kind of three-layer models based on hyperboxes that are constructed in a sequential way. Such a sequential mechanism inevitably leads to the input order and overlap region problem. In this study, we propose a deep FMNN (DFMNN) based on initialization and optimization operation to overcome these limitations. Initialization operation that can solve the input order problem is to design hyperboxes in a simultaneous way, and side parameters have been proposed to control the size of hyperboxes. Optimization operation that can eliminate overlap region problem is realized by means of deep layers, where the number of layers is immediately determined when the overlap among hyperboxes is eliminated. In the optimization process, each layer consists of three sections, namely, the partition section, combination section, and union section. The partition section aims to divide the hyperboxes into a nonoverlapping hyperbox set and an overlapping hyperbox set. The combination section eliminates the overlap problem of overlapping hyperbox set. The union section obtains the optimized hyperbox set in the current layer. DFMNN is evaluated based on a series of benchmark datasets. A comparative analysis illustrates that the proposed DFMNN model outperforms several models previously reported in the literature.

Application of Artificial Neural Networks in the Process Analytical Technology of Pharmaceutical Manufacturing-a Review

Industry 4.0 has started to transform the manufacturing industries by embracing digitalization, automation, and big data, aiming for interconnected systems, autonomous decisions, and smart factories. Machine learning techniques, such as artificial neural networks (ANN), have emerged as potent tools to address the related computational tasks. These advancements have also reached the pharmaceutical industry, where the Process Analytical Technology (PAT) initiative has already paved the way for the real-time analysis of the processes and the science- and risk-based flexible production. This paper aims to assess the potential of ANNs within the PAT concept to aid the modernization of pharmaceutical manufacturing. The current state of ANNs is systematically reviewed for the most common manufacturing steps of solid pharmaceutical products, and possible research gaps and future directions are identified. In this way, this review could aid the further development of machine learning techniques for pharmaceutical production and eventually contribute to the implementation of intelligent manufacturing lines with automated quality assurance.

Prediction of Mefenamic Acid Crystal Shape by Random Forest Classification

OBJECTIVE

Particle shape can have a significant impact on the bulk properties of materials. This study describes the development and application of machine-learning models to predict the crystal shape of mefenamic acid recrystallized from organic solvents.

METHODS

Crystals were grown in 30 different solvents to establish a dataset comprising solvent molecular descriptors, process conditions and crystal shape. Random forest classification models were trained on this data and assessed for prediction accuracy.

RESULTS

The highest prediction accuracy of crystal shape was 93.5% assessed by fourfold cross-validation. When solvents were sequentially excluded from the training data, 32 out of 84 models predicted the shape of mefenamic acid crystals for the excluded solvent with 100% accuracy and a further 21 models had prediction accuracies from 50-100%. Reducing the feature set to only solvent physical property descriptors and supersaturations resulted in higher overall prediction accuracies than the models trained using all available or another selected subset of molecular descriptors. For the 8 solvents on which the models performed poorly (< 50% accuracy), further characterisation of crystals grown in these solvents resulted in the discovery of a new mefenamic acid solvate whereas all other crystals were the previously known form I.

CONCLUSIONS

Random forest classification models using solvent physical property descriptors can reliably predict crystal morphologies for mefenamic acid crystals grown in 20 out of the 28 solvents included in this work. Poor prediction accuracies for the remaining 8 solvents indicate that further factors will be required in the feature set to provide a more generalized predictive morphology model.

Applications of Machine Learning in Solid Oral Dosage Form Development

This review comprehensively summarizes the application of machine learning in solid oral dosage form development over the past three decades. In both academia and industry, machine learning is increasingly applied for multiple preformulation/formulation and process development studies. Further, this review provides the authors' perspectives on how pharmaceutical scientists can use machine learning for right projects and in right ways; some key ingredients include (1) the determination of inputs, outputs, and objectives; (2) the generation of a database containing high-quality data; (3) the development of machine learning models based on dataset training and model optimization; (4) the application of trained models in making predictions for new samples. It is expected by the authors and others that machine learning will promisingly play a more important role in tomorrow's projects for solid oral dosage form development.

Pharmaceutical application of multivariate modelling techniques: a review on the manufacturing of tablets

The tablet manufacturing process is a complex system, especially in continuous manufacturing (CM). It includes multiple unit operations, such as mixing, granulation, and tableting. In tablet manufacturing, critical quality attributes are influenced by multiple factorial relationships between material properties, process variables, and interactions. Moreover, the variation in raw material attributes and manufacturing processes is an inherent characteristic and seriously affects the quality of pharmaceutical products. To deepen our understanding of the tablet manufacturing process, multivariable modeling techniques can replace univariate analysis to investigate tablet manufacturing. In this review, the roles of the most prominent multivariate modeling techniques in the tablet manufacturing process are discussed. The review mainly focuses on applying multivariate modeling techniques to process understanding, optimization, process monitoring, and process control within multiple unit operations. To minimize the errors in the process of modeling, good modeling practice (GMoP) was introduced into the pharmaceutical process. Furthermore, current progress in the continuous manufacturing of tablets and the role of multivariate modeling techniques in continuous manufacturing are introduced. In this review, information is provided to both researchers and manufacturers to improve tablet quality.

Data-smart machine learning methods for predicting composition-dependent Young's modulus of pharmaceutical compacts

The ability to predict mechanical properties of compacted powder blends of Active Pharmaceutical Ingredients (API) and excipients solely from component properties can reduce the amount of 'trial-and-error' involved in formulation design. Machine Learning (ML) can reduce model development time and effort with the imperative of adequate historical data. This work describes the utility of linear and nonlinear ML models for predicting Young's modulus (YM) of directly compressed blends of known excipients and unknown API mixed at arbitrary compositions given only the true density of the API. The models were trained with data from compacts of three BCS Class I APIs and two excipients blended at four drug loadings, three excipient compositions, and compacted to five nominal solid fractions. The prediction accuracy of the models was measured using three cross-validation (CV) schemes. Finally, we demonstrate an application of the model to enable Quality-by-Design in formulation design. Limitations of the models and future work have also been discussed.

Data-Driven Modelling of a Pelleting Process and Prediction of Pellet Physical Properties

In the manufacture of pelleted catalyst products, controlling physical properties of the pellets and limiting their variability is of critical importance. To achieve tight control over these critical quality attributes (CQAs), it is necessary to understand their relationship with the properties of the powder feed and the pelleting process parameters (PPs). This work explores the latter, using standard multivariate methods to gain a better understanding of the sources of process variability and the impact of PPs on the density and strength of the resulting pellets. A Kilian STYL’ONE EVO Compaction Simulator machine was used to produce over 1000 pellets, whose properties were measured, with varied powder feed mechanism and powder feed rate. Process data recorded by the Compaction Simulator machine were analysed using Principal Component Analysis (PCA) to understand the key aspects of variability in the process. This was followed by Partial Least Squares (PLS) regression to predict pellet density and hardness from the Compaction Simulator data. Pellet density was predicted accurately, achieving an R2 metric of 0.87 in 10-fold cross-validation, and 0.86 in an independent hold-out test. Pellet hardness proved more difficult to predict accurately, with an R2 of 0.67 in 10-fold cross-validation, and 0.63 in an independent hold-out test. This may however simply be highlighting measurement quality issues in pellet hardness data. The PLS models provided direct insights into the relationships between pelleting PPs and pellet CQAs and highlighted the potential for such models in process monitoring and control applications. Furthermore, the overall modelling process boosted understanding of the key sources of process and product variability, which can guide future efforts to improve pelleting performance.

A real-time optimization method for economic and effective operation of electrostatic precipitators

Electrostatic precipitators (ESP) have been considered as the main particulate matter (PM) removal facility in the energy industry. This paper presents a real-time optimization method for a one-chamber industrial ESP in an ultra-low emission power plant with an intelligent optimization system (IOS). The IOS seeks to optimize the energy consumption of ESP subject to the outlet concentration requirement in real-time. A coordination control logic is designed to regulate the optimized operation set points with varying operation conditions. The operation optimized by the IOS is compared with the operations under PID (proportion-integral-derivative) and manual control. The results show that the IOS improves the emission compliance rate from 95% of manual control to 100% and the medium concentration is tuned to be 46.6% closer to the emission target. Furthermore, a good balance between emission and energy consumption is achieved, with 35.50% energy conservation for the same emission upper limit of 30 mg/m³. These results prove that the IOS significantly contributes to the efficient operation and economic PM removal by ESP for the energy industry.

IMPLICATIONS

Electrostatic precipitators (ESP) is one of the main PM removal facilities in coal-fired power plants. An intelligent optimization system (IOS) with prediction, optimization, and control modules is designed and constructed for the ESP in an ultra-low emission power plant. A PM removal model is used to predict the outlet concentration of the ESP. The optimal energy consumption of ESP subject to the outlet concentration requirement problem is solved by the particle swarm optimization. A closed-loop and rapping tolerant method is used to eliminate the fluctuation in time-averaged concentration. The system raised is able to ensure the compliance rate while decreasing the energy consumption of the ESP.

Integrating QbD Tools for Flexible Scale-Up Batch Size Selection for Solid Dosage Forms

The pilot scale batch size for solid oral dosage forms is currently defined by major regulatory agencies as one-tenth of the full production, or 100,000 units, whichever is larger. The current criterion is arbitrary and is not based on scientific and risk assessment principles. The approach does not consider geometric, kinematic, and dynamic changes that come into play on scale-up. Even if this criterion is met, impact of scale-up on critical quality attributes cannot be ruled out and also reproducibility cannot be assured simply by restricting the scale-up size. In keeping with the vision for the 21st Century Good Manufacturing Practice initiative to build quality into the product, it is imperative that the selection of scale-up batch size be based on science and risk assessment principles and be part of the product development program. Scale-up should never be seen as an isolated activity. This article will review various tools that can be integrated with quality by design for flexible batch size selection during scale-up.

Chaotic Multi-Objective Particle Swarm Optimization Algorithm Incorporating Clone Immunity

It is generally known that the balance between convergence and diversity is a key issue for solving multi-objective optimization problems. Thus, a chaotic multi-objective particle swarm optimization approach incorporating clone immunity (CICMOPSO) is proposed in this paper. First, points in a non-dominated solution set are mapped to a parallel-cell coordinate system. Then, the status of the particles is evaluated by the Pareto entropy and difference entropy. At the same time, the algorithm parameters are adjusted by feedback information. At the late stage of the algorithm, the local-search ability of the particle swarm still needs to be improved. Logistic mapping and the neighboring immune operator are used to maintain and change the external archive. Experimental test results show that the convergence and diversity of the algorithm are improved.