Multiplex
detection of viable foodborne pathogens is critical for
food safety and public health, yet current assays suffer trade-offs
between cost, assay complexity, sensitivities, and the specificity
between live and dead bacteria. We herein developed a sensing method
using artificial intelligence transcoding (SMART) for rapid, sensitive,
and multiplex profiling of foodborne pathogens. The assay utilizes
the programmable polystyrene (PS) microspheres to encode different
pathogens, inducing subsequent visible signals under conventional
microscopy that can be analyzed using a customized, artificial intelligence-computer
vision, which was trained to decode the intrinsic properties of PS
microspheres to reveal the numbers and types of pathogens. Our approach
enabled the rapid and simultaneous detection of multiple bacteria
from egg samples of <102 CFU/mL without DNA amplification
and showed strong consistency with the standard microbiologic and
genotypic methods. We adopted our assay through phage-guided targeting
to enable the discrimination between live and dead bacteria.