Machine Learning Approach to Enhance the Performance of MNP-Labeled Lateral Flow Immunoassay

• DOI: 10.1007/s40820-019-0239-3
• Publication Date: 2019-01-17T14:11:26Z
• Authors: Wenqiang Yan; Kan Wang; Hao Xu; Xuyang Huo; Qinghui Jin; Daxiang Cui
• Keywords: Multiplex; Point-of-Care Testing; Limiting; Normalization; Point of care

Abstract The use of magnetic nanoparticle (MNP)-labeled immunochromatography test strips (ICTSs) is very important for point-of-care testing (POCT). However, common diagnostic methods cannot accurately analyze the weak signal from ICTSs, limiting applications POCT. In this study, an ultrasensitive multiplex biosensor was designed to overcome limitations capturing and normalization MNPs on ICTSs. A machine learning model sandwich assays constructed used classify weakly positive negative samples, which significantly enhanced specificity sensitivity. potential clinical application evaluated by detecting 50 human chorionic gonadotropin (HCG) samples 59 myocardial infarction serum samples. quantitative range HCG 1–1000 mIU mL −1 ideal detection limit 0.014 , well below threshold. Quantitative results cardiac markers showed good linear correlations with standard values. proposed assay can be readily adapted identifying other biomolecules also in such as environmental monitoring, food analysis, national security.