Machine Learning-enabled optimization of microneedle design for interstitial fluid collection

Utilization of microneedles (MNs) for biological fluid sampling and drug delivery is an emerging field in biotechnology, which contributes greatly to non-invasive methods in medicine. Prior studies on MNs propose designs based on mechanical testing and optimize physical parameters of the MN based on trial-and-error method. While these methods show adequate results, it is possible to enhance the performance of a MN using a large dataset of parameters and their respective performance using advanced data analysis methods. Machine learning (ML) is an important field of study in data analytics that mimics human learning behavior. It is one of the most studied topics in computational sciences which expedites various decision-making processes in many different areas, including biotechnology. In this study, an integration of finite element analysis and ML models are proposed with the purpose of determining the best physical parameters for a MN design, in order to maximize the amount of collected fluid. The fluid behavior in a MN patch is simulated with several different physical and geometrical parameters using COMSOL Multiphysics®, and the resulting dataset is used as the input for ML algorithms including multiple linear regression, random forest regression, decision tree regression, support vector regression, and neural networks. As a comparable method, iterative optimization is performed via LiveLink™ for MATLAB®. This study essentially introduces a novel approach to optimizing MN design for predetermined performance metrics.