THANK YOU FOR SUBSCRIBING
Pharma Tech Outlook | Tuesday, January 19, 2021
Dolomite helps the development of artificial intelligence for microfluidics.
FREMONT, CA: Microfluidic technologies are rapidly becoming the method of choice for micro-and nanoparticle production across several sectors, from cell biology research to drug development. This method provides several benefits over traditional batch methods, including precise process control, improved reliability, and the potential to seamlessly scale up production.
Stay ahead of the industry with exclusive feature stories on the top companies, expert insights and the latest news delivered straight to your inbox. Subscribe today.
Microfluidic particle creation's reproducibility and tunability offer new opportunities to optimize development workflows, using artificial intelligence (AI) to run in silico predictions before laboratory testing. AI and machine learning are now extensively used for several applications in pharmaceutical development. These technologies assist in scaling research and mitigate costs by predicting the experimental parameters or compounds most likely to offer a particular outcome without extensive trial and error requirements. However, AI methods can only be utilized where outcomes are predictable and reproducible, and sufficient curated data are available. Microfluidics brings this precision to particle creation and has enabled users to develop a machine learning model that can successfully forecast the experimental parameters needed to create size-tunable particles with specific characteristics.
Working with multiple data sets offered by Dolomite, the company used a supervised learning approach to train an artificial neural network (ANN) to interpret the non-linear relationships between several experimental parameters – flow rates, polymer concentrations, and many more. – on the resulting particles. Once the company had trained, tested, and validated the ANN model, it used an external data set containing previously unseen data to assess its potential to predict experimental outcomes with accurate and promising outcomes. This proof-of-concept study showed that AI could be coupled with microfluidic systems to scale research. The objective now is to expand this approach to offer researchers a novel tool to scale particle development for several applications.
More in News