Microfluidic devices have revolutionized the the field of cell biology. As microfluidicdevices became affordable and mainstream, they began to serve as miniature labs for detecting and studying biological phenomenon. We develop microfluidic devices for two different biological applications in this thesis. In Chapter 2, we develop a microfluidic device for removal of rare, teratoma-forming cells from human-induced pluripotent stem cell-derived neural progenitors (hiPSC-NPCs). Though generalizable to other cell types, we begin by characterizing the differentiation process of human-induced pluripotent stem cells (hiPSCs) into neural progenitor cells (NPCs) based off residual pluripotency signature and investigate device performance by removing residual teratoma-forming cells. In Chapter 3, we develop a platform for mimicking the early stages of Lyme transmission and apply our device towards detecting Borellia burgdorferi, the causative agent of Lyme Disease. Early research on Lyme disease suggested a model in which the outer surface protein OspA is expressed by Borrelia in the tick, while outer surface protein OspC is expressed in the mammalian host. We first develop a culture system for recreating this dynamic change in surface protein expression, and then demonstrate our device is capable of capturing and detecting Borellia burgdorferi. Teratoma formation remains a safety concern in therapeutic cells derived from hiPSCs. Residual Teratoma forming cells are present in small numbers in differentiated hiPSC cultures and yet are of significant roadblock to the manufacturing and clinical translation of stem cell therapies. Rare cells are often difficult to remove with standard flow cytometry or magnetic bead sorting techniques. Here, we first characterized time-dependent expression of a teratoma marker, stage-specific embryonic antigen 5 (SSEA-5), which binds the H type-1 glycan during neural differentiation of hiPSCs. The fraction of cells SSEA-5+ remained high at 97% on day 3, dropped to 70% on day 4, 40% by day 6, and down to 1% on day 12 of differentiation, indicating successful differentiation. We engineered a microfluidic geometrically enhanced differential immunocapture (GEDI) technology to remove SSEA-5+ rare cells from hiPSC-derived neural progenitor cells (hiPSC-NPCs). The GEDI chip removed more than 95% of teratoma-forming cells and presents a facile tool to potentially functionalize with multiple antibodies and robustly enhance hiPSC-derived cell population safety prior to therapeutic transplantation. The approach is potentially amenable to generate a wide variety of high-quality therapeutic cells and can be integrated within the pipeline of cell manufacturing to improve patient safety and reduce the cost of manufacturing through early removal of undesirable cell types. Lyme Disease is a multisystem infectious disease caused by the Borellia burgdorferi complex, and is a growing threat to public health. Approximately 476,000 people are infected with Lyme in the United States each year. Although Lyme is readily treated with antibiotics when detected early, early detection remains difficult. Current testing remains difficult because the standard 2-tiered ELISA assay indirectly detects Lyme via measurement of a host immune response, which suffers from an inherent time-lag in host antibody production. A direct test for Lyme Disease would overcome these inherent limitations. To this end we report on the first microfluidic immunocapture device for Lyme Disease. We engineered a geometrically enhanced differential immunocapture (GEDI) technology to capture whole-organism Borrelia for direct on-chip detection. This approach is potentially amenable with other work in the field to develop direct PCR tests for Lyme Disease, as our device could serve as a platform to drastically enhance the concentration of present Borrelia into a small volume.