Deep learning-enabled multiplexed point-of-care sensor using a paper-based fluorescence vertical flow assay

Artem Goncharov,Hyou-Arm Joung,Rajesh Ghosh,Gyeo-Re Han,Zachary S. Ballard,Quinn Maloney,Alexandra Bell,Chew Tin Zar Aung,Omai B. Garner,Dino Di Carlo,Aydogan Ozcan
DOI: https://doi.org/10.1002/smll.202300617
2023-01-26
Abstract:We demonstrate multiplexed computational sensing with a point-of-care serodiagnosis assay to simultaneously quantify three biomarkers of acute cardiac injury. This point-of-care sensor includes a paper-based fluorescence vertical flow assay (fxVFA) processed by a low-cost mobile reader, which quantifies the target biomarkers through trained neural networks, all within <15 min of test time using 50 microliters of serum sample per patient. This fxVFA platform is validated using human serum samples to quantify three cardiac biomarkers, i.e., myoglobin, creatine kinase-MB (CK-MB) and heart-type fatty acid binding protein (FABP), achieving less than 0.52 ng/mL limit-of-detection for all three biomarkers with minimal cross-reactivity. Biomarker concentration quantification using the fxVFA that is coupled to neural network-based inference is blindly tested using 46 individually activated cartridges, which showed a high correlation with the ground truth concentrations for all three biomarkers achieving > 0.9 linearity and < 15 % coefficient of variation. The competitive performance of this multiplexed computational fxVFA along with its inexpensive paper-based design and handheld footprint make it a promising point-of-care sensor platform that could expand access to diagnostics in resource-limited settings.
Medical Physics,Applied Physics,Biological Physics
What problem does this paper attempt to address?