Abstract
A functionalized gold nanoparticle-based lateral flow immunoassay (AuNPs-LFIA) platform integrated with a deep learning-based machine vision model for the rapid and quantitative detection of cardiac biomarkers, myoglobin (Myo) and cardiac troponin I (cTnI). AuNPs@Ab probes with high specificity and stability were prepared. AuNPs were synthesized via the citrate reduction method, functionalized with HS-PEG-COOH, and covalently conjugated to antibodies through EDC/NHS-mediated coupling to yield high specificity and stability. Using this platform, the machine vision-assisted AuNPs-LFIA achieved detection limits of 0.224 ng/mL for Myo and 0.071 ng/mL for cTnI. Quantitative readouts were obtained within 8 min, representing a 46.7% improvement in detection efficiency compared to conventional methods. Analysis of spiked serum samples demonstrated a significant correlation with commercial assay kits (R² > 0.99), further validating the platform’s accuracy and clinical applicability. Overall, this AI-enhanced AuNPs-LFIA platform offers sensitive, rapid, and user-friendly quantification. Its robust scalability provides a practical strategy for point-of-care testing (POCT) in cardiovascular diagnostics, particularly in resource-limited settings. GRAPHICAL ABSTRACT: [Image: see text] SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s00604-026-08054-y.