Timely and sensitive detection of cardiac troponin I (cTnI) is critical for early diagnosis of myocardial infarction, particularly at the point-of-care. Herein, we present a novel colorimetric biosensing platform for high-sensitivity detection of cardiac troponin I (cTnI). The platform integrates magnetic particle (MP) anchored locked aptamers, stabilized by short complementary strands to minimize nonspecific folding and background activation prior to target binding, with hyperbranched hybridization chain reaction (HCR) and catalytic DNA (DNAzyme) nanocomplex-mediated signal amplification. This enzyme-free amplification system detects cTnI directly in 25-30 min, with a calculated detection limit of 0.25 ng L-1, a wide dynamic range of 0.5-50 000 ng L-1, and a coefficient of variation below 5% using just 25 µL of patient serum. The developed assay was evaluated using both human and canine serum samples. To assess classification performance, three distinct hyperparameter optimization strategies were applied to a reduced feature space. The model achieved an accuracy of 90.91% and a recall of 89.89% for human samples, and an accuracy of 83.33% with a recall of 85.71% for canine samples. Blind testing with human serum samples further confirmed the robustness of the platform, showing an overall accuracy of around 90%. This integrated biosensing and machine learning framework enables rapid and sensitive detection of cardiac troponin I, demonstrating strong potential for myocardial infarction diagnosis across species in a pre-clinical setting.
Tripathy et al. (Mon,) studied this question.