New Strategies for High Efficiency and Precision Bioprinting by DOE Technology and Machine Learning

Abstract

Extrusion-based 3D printing technology is currently demonstrating considerable potential in the field of tissue engineering scaffolds, enabling the construction of in vitro models with complex structures and functions using a wide range of biomaterials and cells at a low cost. In recent years, researchers have spent considerable effort developing novel bio-inks and employing a greater variety of cell sources to enhance biological compatibility and functionality. However, the majority of current bio-ink materials are unprintable due to their low viscosity and long curing time, as well as insufficient shape fidelity before the secondary cross-linking process. The study aims to bridge this gap by optimizing the material ratios and predicting the printing process before work. This article presents new strategies for the design, fabrication, and analysis of a new composite bio-ink material. The optimal ink ratios are verified by a design of experiments (DOE) experimental design and evaluation metrics for printing printability (Pr) values. A machine learning model is used to predict the ink printing area and determine the printing process parameters. The influence mechanism of ink materials with different concentrations of poly (ethylene glycol) diacrylate (PEGDA) ratios on printed fibers is investigated. Finally, the optimal results are used as an example to demonstrate the printability of multilayer stents. Thus, the design approach allows for the rapid and cost-effective exploration of novel ink ratios, while also providing higher fidelity and more accurate process metrics for the fabrication of tissue structures with multidimensional variables.