Rapid pathogen ID and AST for bacteremia diagnosis via integration of immunocapture, microfluidics, and artificial intelligence

The National Institute of Allergy and Infectious Diseases awarded the Board of Trustees of Southern Illinois University $547,292 on September 1, 2026, under the Allergy and Infectious Diseases Research program (CFDA 93.855) to develop a rapid diagnostic system for bacteremia pathogen identification and antimicrobial susceptibility testing.

The project integrates immunocapture, microfluidics, and artificial intelligence to detect bacteria directly from blood samples at low concentrations (1–103 CFU/mL), addressing a critical gap in current diagnostic methods that rely on blood culture and require 2–7 days for results. The system will employ magnetic nanoparticles functionalized with immune proteins to capture bacteria while depleting blood matrix components, then pass processed samples through a microfluidic device with integrated filtration to concentrate bacteria in microwells. Pathogen identification will be conducted via intracellular molecular probing of gene markers, and antimicrobial susceptibility testing will monitor bacterial growth in the presence of antibiotics across concentration gradients. A deep-learning artificial intelligence algorithm will automatically analyze single-cell data to deliver real-time, high-accuracy results. The research focuses on the two most prevalent bacteremia-causing species, Escherichia coli and Staphylococcus aureus. By significantly shortening diagnostic turnaround time, the system aims to enable targeted, evidence-based antibiotic stewardship to improve patient survival and reduce antimicrobial resistance.

Performance occurs in Carbondale, Illinois, with a period of performance from September 1, 2026, through August 31, 2029.

Abstract

Bacteremia is a life-threatening condition affecting millions globally, with high mortality rates and substantial healthcare costs. Early and precise pathogen identification (ID) and antimicrobial susceptibility testing (AST) are critical for timely and effective treatment, yet current diagnostic methods rely on blood culture, requiring 2-7 days to deliver results. This delay often forces empirical broad-spectrum antibiotic use, contributing to antimicrobial resistance (AMR) and poor clinical outcomes. This project aims to develop a rapid diagnostic system by integrating immunocapture, microfluidics, and artificial intelligence (AI) to enable early, evidence-based antibiotic stewardship. The proposed system is designed to directly detect bacteria at low concentrations (1-103 CFU/ML) from blood samples, a key challenge in bacteremia diagnosis. Magnetic nanoparticles (mnps) functionalized with immune proteins will be employed to capture bacteria while depleting the blood matrix. The processed sample will then pass through a microfluidic device with integrated filtration to concentrate bacteria in microwells, enabling single-cell ID and AST. Pathogen ID will be conducted via intracellular molecular probing of gene markers, and AST will be performed by monitoring bacterial growth in the presence of antibiotics across concentration gradients. The system's deep-learning AI algorithm will automatically analyze single-cell data, providing real-time, high-accuracy pathogen ID and AST results. By significantly shortening the time required for bacteremia diagnosis, this study aims to provide targeted, evidence-based antibiotic stewardship to improve patient survival, reduce AMR, and ultimately transform bacteremia management. This research will focus on the two most prevalent bacteremia-causing species, escherichia coli and staphylococcus aureus, and evaluate three commonly used antibiotics. The proposed method will be validated against the standard microdilution approach, ensuring clinical relevance. Additionally, potential limitations such as blood matrix interference and bacterial detection sensitivity will be systematically addressed. Beyond its scientific contributions, this project will enhance undergraduate and graduate training in biomedical engineering, microbiology, and AI-driven diagnostics. Research findings will be incorporated into university courses, benefiting approximately 70 undergraduates annually.