Combining Ultralong Organic Phosphorescent Nanoparticle-Based Lateral Flow Strip Assay With Deep Learning: A Quantitative Detection Method for Creatine Kinase MB

In recent years, there has been an increase in the number of cases of acute chest pain, especially those caused by acute myocardial infarction (AMI), which poses a significant health risk. Early and accurate diagnosis of AMI is essential to improve patient prognosis. We used the amphiphilic surfactant polyvinylpyrrolidone (PVP) to encapsulate high-efficiency organic phosphor molecules (with a phosphorescent quantum yield of 38%) and used such ultralong organic phosphorescent nanoparticles as labels for the lateral flow system to develop a lateral flow system for creatine kinase MB isoenzyme (CK-MB) quantification. We also invoked deep learning to construct a lightweight network model based on residual modules to achieve quantitative detection. We used data augmentation (DA) to provide the possibility of a high-precision detection model based on limited samples. Our proposed detection method can detect CK-MB solutions at concentrations of 0.3–100 ng/ml and has the advantages of lower cost, less sample requirement, shorter detection time, and more straightforward instrumentation requirements than traditional methods, contributing to more efficient clinical decision-making and treatment.