BMF meeting covers AI in biomanufacturing

AI is rapidly transforming health care, offering immense opportunities for cost savings and expansion in the pharmaceutical industry, Bruker said.  

Large pharma companies are partnering with AI providers to explore drug discovery, clinical research, and overall capability enhancement. Kymanox has established a multidisciplinary task force to explore internal and external opportunities in AI to better serve clients, he said. 

Bruker said that the FDA is actively involved in developing regulations that promote AI use while addressing potential risks. The agency has published several guidance documents, including the "Good Machine Learning Practice" and "Marketing Submission Recommendations for Artificial Intelligence and Machine Learning Algorithms in Software as a Medical Device." 

These documents outline the FDA's expectations for AI development and deployment in health care. They emphasize the importance of human oversight, data quality and ongoing performance monitoring. 

The FDA is also encouraging manufacturers to develop and submit predetermined change control plans for AI-based medical devices, Burker said. PCCPs should describe the process for making changes to AI models after initial approval. 

One example of an FDA-approved AI-based medical device that Bruker gave is Medtronics’ GI Genius, a visualization system that helps surgeons identify polyps during colonoscopies. GI Genius has been shown to improve adenoma detection rates compared to traditional colonoscopy. 

The FDA emphasizes that AI should be used under human control, with companies bearing responsibility for its outputs, Burker said. The agency is actively involved in developing regulations that promote AI utilization while addressing potential risks. 

It is important to use AI as intended and understand its pitfalls. AI should not replace human decision-making, and it is important to monitor its performance over time, Bruker said. 

Monteiro discussed how biopharma can benefit from generative analytics and data storytelling.  

• Generative analytics is a new field that uses artificial intelligence to analyze data and provide insights to users.  
• Data storytelling is the process of using data to tell a compelling story that can be used to make decisions. 

Generative analytics can help biopharma companies save time and money by automating tasks and providing insights that can be used to make better decisions, Monteiro said. For example, generative analytics can be used to identify patterns and trends in data that can be used to improve pricing, quality management, risk management and environmental, social and corporate governance. 

Data storytelling can help biopharma companies communicate the insights that they have gained from data in a way that is easy to understand and actionable. This can help to improve decision-making and collaboration within biopharma companies, she said. 

In addition to the benefits discussed in the presentation, she said that generative analytics and data storytelling can also help biopharma companies to 

• improve compliance with regulations, 
• develop new products and services and 
• improve patient care. 

Overall, generative analytics and data storytelling are powerful tools that can help biopharma companies to achieve their strategic goals, Monteiro said. 

Metabolic flux analysis is a technique used to measure and monitor the rates of chemical reactions in a biological system. MFA can be used to optimize bioreactor processes, select cell lines and predict what will happen if you change something in the bioreactor, Yenne said. 

MFA is based on the idea that the rates of chemical reactions in a biological system can be determined by measuring the concentrations of metabolites (the products and reactants of the reactions) at different times. This information can then be used to construct a mathematical model of the system that can be used to predict the behavior of the system under different conditions. 

MFA has been used successfully to optimize a variety of bioreactor processes, including the production of pharmaceuticals, biofuels and food additives, Yenne said. MFA can also be used to select cell lines that are more productive or have other desirable properties. Additionally, MFA can be used to predict what will happen if you change something in the bioreactor, such as the amount of raw material that is added or the temperature of the reactor. 

MFA is a powerful tool that can be used to improve the efficiency and productivity of bioreactor processes, he said. However, it is important to note that MFA is a complex technique that requires specialized expertise to implement.