Machine Learning-Assistant Colorimetric Sensor Arrays for Intelligent and Rapid Diagnosis of Urinary Tract Infection
Jianyu Yang, Ge Li, Shihong Chen, Xiaozhi Su, Dong Xu, Yueming Zhai, Yuhang Liu, Guangxuan Hu, Chunxian Guo, Hong Bin Yang, Luigi G. Occhipinti, Fang Hu
Abstract
Urinary tract infections (UTIs), which can lead to pyelonephritis, urosepsis, and even death, are among the most prevalent infectious diseases worldwide, with a notable increase in treatment costs due to the emergence of drug-resistant pathogens. Current diagnostic strategies for UTIs, such as urine culture and flow cytometry, require time-consuming protocols and expensive equipment. We present here a machine learning-assisted colorimetric sensor array based on recognition of ligand-functionalized Fe single-atom nanozymes (SANs) for the identification of microorganisms at the order, genus, and species levels. Colorimetric sensor arrays are built from the SAN Fe 1 –NC functionalized with four types of recognition ligands, generating unique microbial identification fingerprints. By integrating the colorimetric sensor arrays with a trained computational classification model, the platform can identify more than 10 microorganisms in UTI urine samples within 1 h. Diagnostic accuracy of up to 97% was achieved in 60 UTI clinical samples, holding great potential for translation into clinical practice applications.