Development of Rapid and Highly Sensitive Microfluidic and Spectroscopic Techniques for Clinical Diagnosis and Environmental Detection

Microfluidics is a field of study that fabricates small channels into substrates to conduct research in multiple areas. Point-of-care (PoC) diagnostics, nanomaterial reactions, and lab-on-a-chip flourish in microfluidics due to the small amount of reagent or clinical sample needed. Channels can be 3D printed or lithographed onto silicon wafers depending on the size of the channel needed, biocompatibility, and design flexibility. With 3D-printed microfluidics becoming more popular, microfluidic chips have an advantage in being low cost, high uniformity, high sensitivity, and high specificity for disease diagnostics. Sepsis is a life-threatening immune response to infection in the body, eventually resulting in fatal organ failure. In Chapter 2, we present a 3D-printed microfluidic chip that bioconjugates antibodies CD69, CD64, and CD25 to channel surfaces to capture sepsis cells in blood samples and validate it with clinical samples (n = 125 septic, n = 10 healthy). Statistical differences were also found between healthy volunteer and sepsis patient antigen cell counts (CD69 p-value < 0.001, CD64 p-value < 0.004, CD25 p-value < 0.0009), and primarily using antigen expression counts were confirmed using principal component analysis. The receiver operating characteristic (ROC) analysis showed an area under the curve (AUC) of 0.989, 0.988, and 0.992 for CD69, CD64, and CD25, respectively, and a combined biomarker panel of 0.997. Urinary tract infections are among the most common infections present in women, as well as in geriatric adults. If untreated, UTIs can lead to urosepsis and death, especially in more vulnerable population. In Chapter 3, we present a new microfluidic device for rapid precise analysis of white blood cells found in urine, with a limit of detection down to 3 cells/µL. We verify the device for future use utilizing antibody mismatch and urinary crystals, along with methods that will shape the clinical use of the microfluidic chip. High Frequency Transmission Spectroscopy also has potential to be used for microfluidics, as introduced in Chapter 4. It is a novel technique wherein electromagnetic waves propagate through an aqueous solution to obtain an electrical signal. Although similar to impedance, HFTS utilizes an ENA Network Analyzer to measure analytes via their transmitted potential (voltage in/voltage out), and is electrically isolated from the analyte of interest. Its full potential is unknown, but it can be used to differentiate electrolytes at varying concentrations by integrating machine learning methods. With time, microfluidic devices can also be integrated into HFTS for future work in biochemical areas such as cells, tissues, and nanomaterials. As showcased in Chapter 5, this technique has high potential for determining environmental contaminants, such as the presence of radioisotopes and other toxic chemical species that can be found in aqueous domains. SrCl2 was measured from 10-2 to 10-10 M and was analyzed alongside other electrolytes to determine if they can identified and differentiated from one another using machine learning. Supervised classification of SrCl2 concentrations achieved testing accuracies of 99% to 100% and electrolyte discrimination from 98% to 100% accuracy. These results demonstrate the feasibility of HFTS as a method for material classification in aqueous solutions at low concentrations. In this dissertation, we present advancements in microfluidic devices for disease diagnostics, as well as its use in configuring HFTS. Future studies will expand on these findings for microfluidics and introduce new concepts for HFTS.