Date of Award

12-2022

Degree Type

Dissertation

Degree Name

Doctor of Philosophy in Engineering

Department

Electrical Engineering and Computer Science

First Advisor

Siu-Tang Yau

Second Advisor

Lili Dong

Third Advisor

Hongkai Yu

Abstract

The current methods of the diagnosis of bloodstream infections are based on bacterial culture growth, a process that requires considerable time, e.g., 12-16 hours, to obtain a result. This long wait time for the result creates many problems, including the generation of multi-drug resistant organisms (MDROs). At the same time, infected bloodstream usually contains a very low concentration of bacteria, i.e., lower than 5 CFU/mL. The long diagnosis time and the extremely low concentration of bacteria in the infected bloodstream make such infections difficult to diagnose. Here, we demonstrate a culture-free approach for the diagnosis of bloodstream infections using a modified immunoassay-based detection platform. Using this platform to detect bacteria in blood samples will not require culture and therefore can significantly shorten the diagnosis time from 12-16 hours to about 2 hours. This dissertation first presents a feasibility study of using the platform in the diagnosis of bacterial infections. This study describes the operation and performance of the platform on prepared blood samples. The results of using the platform to perform the three-step diagnosis, namely detection, identification, and antibiotic susceptibility testing (AST) are presented and analyzed in detail. The major part of the dissertation describes the results of using the platform to perform the three-step diagnosis on a cohort of clinical blood samples. The clinical samples include whole blood samples and blood samples spiked with bacteria isolated from clinical samples. The diagnosis results obtained using the platform was compared with those obtained using state-of-the-art technologies. The results suggest the game-changing nature of the platform in the diagnosis of infections.

Download
COinS