Expeditious and accurate determination of pathogenic bacteria cell viability is of great importance to public health for numerous areas including medical diagnostics, food safety, and environmental monitoring. In this work a cell buoyant mass classifier approach is presented to assess bacteria cell viability in real time. Buoyant mass measurements for live and dead Gram-positive and Gram-negative bacteria populations were acquired with a commercial suspended microchannel resonator, Archimedes, to generate receiver operating characteristic (ROC) curves. To quantitatively assess the difference in buoyant mass for live and dead bacteria populations, ROC curves were generated to demonstrate cell viability determination. The results are presented as a binary classifier with a decision boundary, above which cells are considered live and below which cells are considered dead. A decision threshold value is evaluated with consideration that a certain true positive rate (correct classification of a live cell) is maintained with an acceptable false positive rate. The potential for this approach to monitor cell viability in real time is significant, especially when considering multiple classifier dimensions such as buoyant mass and density. This classifier approach represents a next generation technique for rapid and label-free diagnostics based on cell feature measurements.
© 2022. This is a U.S. government work and not under copyright protection in the U.S.; foreign copyright protection may apply.