Facile Quantification of Nanosized Bioparticles in Bright-field Micrograph via Deep Learning

The pathogen is considered a critical threat to public health, especially immunocompromised people such as babies, and the elderly. Initiating the apt therapy to the defected patients before turning into the peak viral load is significant to prevent their fatal damage. However, most point-of-care diagnostics are unsuitable for counting the concentration of the virus. Accordingly, various bioimaging techniques are suggested to visualize and/or count biological samples like histopathology. Despite their superb resolution and precise quantitation, the need for sophisticated equipment or professional manpower stunts the broad interest.<br/>Explosive interest in artificial intelligence (AI) induces profuse interdisciplinary studies, especially in biomedical engineering and clinical medicine recently. Due to the property of pure convolution and superior object categorization skills, convolutional neural networks (CNNs) that model human vision has been demonstrated in comprehensive clinical data to predict patient diagnoses. Even though CNNs produce high capabilities for the assigned task, plenty of pre-processing and the requirement of expert interpretation are still mandated.<br/>Tri-layer resonator is adopted as the optical solution to break through weak light-matter interaction in far-field optics. By harnessing thin-film interference, the modulated light produces a strong resonance in the bioparticles of low refractive index and nanoscale size and leads to the perception of targeted analytes. Additionally, the aggregation of biofunctionalized nanoparticles in drop-casted solution due to hydrodynamics of an evaporating droplet (e.g., the Marangoni flow) synergizes with the slow-light effect so that the invisible analytes can be discovered with definite chromatic information even through the usual optical microscope (OM). During the nanofabrication and surface-functionalizing, however, the residues and defects randomly exist on the surface of the designed biosensor. Hence, we introduce the CNN that establishes a correlation between the chromatic information of optical micrographs and the ground-truth of corresponding scanning electron microscope (SEM) images.<br/>Here, we showcase the chromatic immunoassay system consisting of the biosensor that is optically optimized for general zoonotic virus and deep learning that is robustly constructed with 1596 pairs of OM images and matching SEM images. The hard negative samples of the fabrication fluctuation including the impurities and the defects are trained to identify the desired particles automatically, thus averting the false-positive and enabling accurate quantitative analysis. The image-driven biosensing system validates the limit of viral detection of 104 copies/mL which is lower than the rapid antigen test using lateral flow immunoassay (LFIA) and achieves high sensitivity and specificity for the diverse bioparticles modeling Zika virus, Monkeypox virus, and Mumps virus. The presented immunoassay platform may also be amenable to microscopic hazardous fragments such as metal oxide nanoparticles and microplastics.<br/>In this study, we have shown that CNN-based bright-field micrograph analysis allows not only intuitive immunoassay but also quantification of minuscule subjects. Nanosized bioparticles are detected with a straightforward antibody-antigen reaction, which does not require either labeling or amplification. A single drop of the solution enables visceral recognition through the vision. Owing to the purity of the CNN, the various analytes whose diameters are below the diffraction limit can be estimated with the little training set. We also have confirmed that the presented immunoassay quantifies the concentration accurately within the average viral load from SARS-CoV-2-infected patients. We believe that the streamlined quantitative system is applicable to numerous nanoscale dicey particles.