Machine-learning-integrated NLRC3 and cardiac troponin T dual-analyte lateral-flow biosensor for point-of-care acute coronary syndrome screening and acute myocardial infarction-unstable angina discrimination.

The implementation of high-sensitivity cardiac troponin assays has significantly improved the prompt diagnosis of acute myocardial infarction (AMI). Nevertheless, early risk stratification and the swift diagnosis of acute coronary syndromes (ACS) remain challenging. We report a machine-learning-integrated NLRC3/cTnT dual-analyte time-resolved fluorescent lateral-flow immunoassay biosensor for quantitative ACS screening and discrimination of acute myocardial infarction (AMI) from unstable angina (UA). The biosensor uses europium nanoparticle time-resolved fluorescence in a dual-channel lateral-flow setup to simultaneously detect NLRC3 and cTnT within 20 min. The original serum sample is diluted separately for each channel, with 100 µL of the diluted sample applied to each. The optimized NLRC3 and cTnT assays demonstrated strong quantitative performance, precision, minimal cross-reactivity, and acceptable interference resistance. In clinical evaluation, NLRC3 alone identified ACS with 80.5% accuracy but was less effective in distinguishing AMI-UA. By combining NLRC3 and cTnT using a Naïve Bayes classifier, the dual-analyte biosensor improved to 84.6% sensitivity, 89.6% specificity, and 87.8% accuracy for AMI-UA discrimination. This study introduces a dual-pathophysiology lateral-flow biosensor that merges inflammatory-risk and myocardial-injury detection, providing a practical POCT method for decentralized ACS triage.