The future of subvisible particle and protein aggregate image analysis is here – are you ready?
ParticleSentryᴬᴵ software delivers the future of image analysis to biopharmaceutical manufacturing. Whether you’re just considering moving forward with artificial intelligence, or your internal machine learning project hasn’t delivered the results you had hoped, patent-pending ParticleSentryᴬᴵ software can help – read on for valuable information to inform your project.
What is the difference between Artificial Intelligence, Machine Learning and Deep Machine Learning?
Machine Learning (ML). Deep Machine Learning (DML). Artificial Intelligence (AI). There isn’t a day that passes that we’re not hearing about advancements in our world brought forward by AI or ML. The terms are often used synonymously, but they are different.
AI is at best poorly defined. Experts tend to agree that AI is when a computer can emulate human behaviors like problem-solving, planning, learning complex functions, understanding, and seeking patterns, etc. AI is a larger concept that allows for machine learning.
ML is a form of artificial intelligence that enables machines (computers, etc.) to learn by analyzing patterns that humans cannot interpret or would have a hard time doing so. An excellent example of this is the large data sets that computers can analyze, which would fatigue a human. Computers can run the exact same process on vast data sets over, and over and over. ML is task-focused and is used to make predictions or answer questions. ML is an enabler of many of the features in online recommendations, search algorithms, email filtering, etc.
DML is considered a subset, or an “evolution of machine learning”  and uses algorithms to construct neural networks. Convolutional neural networks (ConvNets or CNNs) typically support classification and computer vision vision tasks .
Which of these applies to ParticleSentryᴬᴵ?
ParticleSentryᴬᴵ software employs DML in the visual analysis of subvisible particle (SbVP) images captured from particle imaging platforms like backgrounded membrane imaging, flow imaging microscopy, flow cytometry, and microfluidic imaging. Because DML is a subset of ML, and ML is a part of AI, all three apply.
Why isn’t subject matter expert (SME) analysis of images or basic software assisted SbVP image classification an effective solution?
When we meet with customers to discuss our technology, we are often asked if we can parse particle images based on a qualitative set of human interpretable features – like silicone oil, air bubbles, etc. Even an SME can only visually resolve gross features in the amorphous, often blurry images. We could easily deliver a solution that classifies these images by human defined type – but it would likely only tell you what you already know: that your sample contains air bubbles, silicone oil droplets, SbVPs that are likely protein aggregates, silicone oil + protein particles and aggregates. These image classifiers are only as good as the data fed to teach the classifiers. When they encounter an unknown particle, it is forced into its nearest classification category. (Figure 1)
SbVPs all have morphological and textural features that tell stories about their origins, different drug formulations and stress conditions result in different features. Those tiny feature differences are virtually impossible for humans to view and characterize. Even if SME is very good they can’t reliably see below 50 µm, let alone discern the embedded subtleties contained in the image. 50 µm is an order of magnitude larger than particle sizes of concern in a biologic drug product today. If your SMEs create defect libraries based on their qualitative analysis of these images, use of the library is limited by human perception of those features. Add fatigue and variations in SME perceptions, and grouped populations are even more qualitative, virtually ensuring validation of the method is unachievable. Further, these libraries are limited to the SbVPs that your SMEs have seen and classified. If you employ an ML classification algorithm to these libraries for analysis of particles from other samples, the particle you’ve never seen or classified is grouped into the nearest class match – which may not be a match. These simple classification solutions, whether human or software assisted, even when based upon quantitative data like aspect ratio, circularity, or longest dimension, must put unknown SbVPs into a class – even when it isn’t a match. These ML classifiers are task focused on forcing a particle image into a known, labeled class, not designed to watch for the unique, new feature or feature variations, or new outlier particle classes.
Our approach looks at the heterogeneous mix of particles (aggregates, SbVPs, air/gas bubbles, silicone oil) in a sample. By employing well-designed experiments and looking at the entire sample, we simplify sample data preparation, and deliver statistical tools to identify shifts in any of these constituent populations, or identify outlier particle populations not previously seen. There is definitely a time in drug development to understand exactly what each particle is, its origin and risk – and those tests will be discussed at length in the next article in this series.
The parsing question stems from wanting to see and understand every SbVP in every sample, rather than using statistical methods to detect a shift in the normal particle population. We call this method “seeking the needle in the haystack”. This approach doesn’t challenge that old identify-every-particle-paradigm or leverage the rich textural and morphological features contained in the images that are not detectable by human SMEs. DML allows us to leverage these features and still identify statistical variations in particle features – the new features or characteristics associated with outlier particles. Our process quantifies all the particles in your sample – the entire heterogeneous mix of silicone oil bubbles, air/gas bubbles, protein aggregates, SbVP’s, etc. Instead of classifying all the needles in the particle haystack, we characterize the haystack. Our approach allows us to deliver quantitative data and demonstrate when a statistical shift has occurred in your haystack or particle population. (Figure 2)
We want to implement AI/ML/DML within our analytical team – where do we start?
While everyone wants to jump on the AI bandwagon, not all problems are appropriate for these often-complex applications. Regularly we hear about ML efforts that are launched within the analytical teams at our biopharmaceutical customers. Most of those programs have not gained traction because they were either too large in scope or driven by the belief that AI/ML will solve a poorly defined problem – the classification examples above are characteristic of what we see.
Our advice as a software company with a DML product…
1. Not leveraging the morphological and textural detail in the SbVP particle images
from imaging platform(s).
2. The library of particle images requires SME intervention to identify new particle
3. Quantitative methods are needed to allow comparison of particle populations
between production lines and at different sites.
4. Methods are needed to compare process-to-process that allow the capture of
morphological changes and provide quantitative and statistical tools to interpret
Why ParticleSentryᴬᴵ software?
Regardless of when you head down the ML path, make sure that you’re working with a partner that understands your challenges, has knowledge of the industry and is agile. We hope that partner is SentrySciences, let’s talk about how we can work together.
1Deep learning vs. machine learning: What’s the difference? https://www.zendesk.com/blog/machine-learning-and-deep-learning/11/7/22
2Convolutional Neural Networks. IBM Cloud Education. https://www.ibm.com/cloud/learn/convolutional-neural-networks%2011/7/22.
Difference between Artificial Intelligence and Machine Learning.
Difference Between Machine Learning and Artificial Intelligence. https://www.geeksforgeeks.org/difference-between-machine-learning-and-artificial-intelligence/11/3/22.
Deep learning vs. machine learning in Azure Machine Learning. https://learn.microsoft.com/en-us/azure/machine-learning/concept-deep-learning-vs-machine-learning 11/7/22.
Machine Learning. IBM Cloud Education. https://www.ibm.com/cloud/learn/machine-learning 11/7/22.
Calderon CP, Daniels AL, Randolph TW. Deep Convolutional Neural Network Analysis of Flow Imaging Microscopy Data to Classify Subvisible Particles in Protein Formulations. J Pharm Sci 2018; 107(4) 999-1008. https://doi.org/10.1016/j.xphs.2017.12.008.
Sign up to receive our quarterly newsletters