AI-Augmented 3dp microfluidic bio-reaction system for point-of-care detection of mpox via LAMP-on-a-chip

Microfluidic devices hold enormous potential for high-throughput, low-cost, and point-of-care (PoC) biological assays, yet current fabrication and control strategies limit their scalability and automation. In this thesis, we first demonstrate that digital light processing (DLP) 3D printing provides a rapid and economical alternative to conventional soft lithography by reproducing a complex organ-on-chip. We then integrate finite element modeling (FEM) simulations with machine learning (ML) tools to create a user-friendly interface for designing single-phase flow circuits that achieve target flow rates with high accuracy. Next, we introduce programmable 3D microfluidic reaction reservoirs that utilize pressure-dependent two-phase (water/air) fluid control in hydrophobic channels. Adjusting the inlet pressure allows fluid to be temporarily retained in these reservoirs for precisely timed reactions, and then released into following reservoirs, facilitating automated consecutive or parallel assays. To further expand accessibility, a low-cost, portable pressure pump was developed to eliminate reliance on bulky, expensive equipment. The utility of the integrated approach is demonstrated through a PoC Loop-mediated Isothermal Amplification (LAMP)-on-chip assay targeting monkeypox markers. The sample fluid, heated to 65°C for 30 minutes, interacts with pre-immobilized LAMP primers within the reservoir, producing a distinct colorimetric shift that confirms the presence of the pathogen. The automated control of fluid handling and reaction timing proved the system's potential for PoC diagnostics. Collectively, the presented methodologies pave the way for more accessible, precise, and autonomous PoC bioassays.