Towards quantification and differentiation of protein aggregates and silicone oil droplets in the low micrometer and submicrometer size range by using oil-immersion flow imaging microscopy and convolutional neural networks

Sub-Visible Particle Classification and Label Consistency Analysis for Flow-Imaging Microscopy Via Machine Learning Methods

Sub-visible particles can be a quality concern in pharmaceutical products, especially parenteral preparations. To quantify and characterize these particles, liquid samples may be passed through a flow-imaging microscopy instrument that also generates images of each detected particle. Machine learning techniques have increasingly been applied to this kind of data to detect changes in experimental conditions or classify specific types of particles, primarily focusing on silicone oil. That technique generally requires manual labeling of particle images by subject matter experts, a time-consuming and complex task. In this study, we created artificial datasets of silicone oil, protein particles, and glass particles that mimicked complex datasets of particles found in biopharmaceutical products. We used unsupervised learning techniques to effectively describe particle composition by sample. We then trained independent one-class classifiers to detect specific particle populations: silicone oil and glass particles. We also studied the consistency of the particle labels used to evaluate these models. Our results show that one-class classifiers are a reasonable choice for handling heterogeneous flow-imaging microscopy data and that unsupervised learning can aid in the labeling process. However, we found agreement among experts to be rather low, especially for smaller particles (< 8 µm for our Micro-Flow Imaging data). Given the fact that particle label confidence is not usually reported in the literature, we recommend more careful assessment of this topic in the future.

Image Classification of Degraded Polysorbate, Protein and Silicone Oil Sub-Visible Particles Detected by Flow-Imaging Microscopy in Biopharmaceuticals Using a Convolutional Neural Network Model

Degradation of polysorbates in biopharmaceutical formulations can induce the formation of sub-visible particles (SvPs) in the form of free-fatty acids (FFAs) and potentially protein aggregates. Flow-imaging microscopy (FIM) is one of the most common techniques for enumerating and characterizing the SvPs, allowing for collection of image data of the SvPs in the size ranges of two to several hundred micrometers. The vast amounts of data obtained with FIM do not allow for rapid manual characterization by an experienced analyst and can be ambiguous. In this work, we present the application of a custom convolutional neural network (CNN) for classification of SvP images of FFAs, proteinaceous particles and silicon oil droplets, by FIM. The network was then used to predict the composition of artificially pooled test samples of unknown and labeled data with varying compositions. Minor misclassifications were observed between the FFAs and proteinaceous particles, considered tolerable for application to pharmaceutical development. The network is considered to be suitable for fast and robust classification of the most common SvPs found during FIM analysis.

Novel machine learning models for flow imaging microscopy sub-visible particle classification in protein formulations

Understanding the particulate content of formulated drug products is essential for ensuring patient safety. In particular, it is critical to assess the presence of aggregated proteins or extraneous particles (e.g. fibres) that pose potential dangers. Additionally, it is useful to be able to distinguish non-proteinaceous particles, such as silicone oil droplets that commonly occur in formulations stored in pre-filled syringes. Standard particle counting methods (e.g. light obscuration) provide only total numbers of particles of a given size, but provide no mechanism for particle classification. Significant recent work has focused on the use of flow imaging microscopy to enable simultaneous classification and counting of particles using machine learning (ML) models including convolutional neural networks (CNN). In this paper we expand upon this theme by exploring techniques for achieving high prediction accuracy when the size of the labeled dataset used for model training is limited. We demonstrate that maximum performance can be achieved by combining multiple techniques such as data augmentation, transfer learning, and novel (to this field) models combining imaging and tabular data.

Transfer Learning Analysis for Subvisible Particle Flow Imaging of Pharmaceutical Formulations

Subvisible particles are an ongoing problem in biotherapeutic injectable pharmaceutical formulations, and their identification is an important prerequisite for tracing them back to their source and optimizing the process. Flow imaging microscopy (FIM) is a favored imaging technique, mainly because of its ability to achieve rapid batch imaging of subvisible particles in solution with excellent imaging quality. This study used VGG16 after transfer learning to identify subvisible particle images acquired using FlowCam. We manually prepared standards for seven classes of particles, acquired the image information through FlowCam, and fed the images over 5 µm into VGG16 consisting of a convolutional base of VGG16 pre-trained with ImageNet data and a custom classifier for training. An accuracy of 97.51% was obtained for the test set data. The study also demonstrated that the recognition method using transfer learning outperforms machine learning methods based on morphological parameters in terms of accuracy, and has a significant training speed advantage over scratch-trained CNN. The combination of transfer learning and FIM images is expected to provide a general and accurate data-analysis method for identifying subvisible particles.

Particles in Biopharmaceutical Formulations, Part 2: An Update on Analytical Techniques and Applications for Therapeutic Proteins, Viruses, Vaccines and Cells

Particles in biopharmaceutical formulations remain a hot topic in drug product development. With new product classes emerging it is crucial to discriminate particulate active pharmaceutical ingredients from particulate impurities. Technical improvements, new analytical developments and emerging tools (e.g., machine learning tools) increase the amount of information generated for particles. For a proper interpretation and judgment of the generated data a thorough understanding of the measurement principle, suitable application fields and potential limitations and pitfalls is required. Our review provides a comprehensive overview of novel particle analysis techniques emerging in the last decade for particulate impurities in therapeutic protein formulations (protein-related, excipient-related and primary packaging material-related), as well as particulate biopharmaceutical formulations (virus particles, virus-like particles, lipid nanoparticles and cell-based medicinal products). In addition, we review the literature on applications, describe specific analytical approaches and illustrate advantages and drawbacks of currently available techniques for particulate biopharmaceutical formulations.