Rapid label-free identification of pathogenic bacteria species from a minute quantity exploiting three-dimensional quantitative phase imaging and artificial neural network

Abstract

For appropriate treatments of infectious diseases, rapid identification of the pathogens is crucial. Here, we developed a rapid and label-free method for identifying common bacterial pathogens as individual bacteria by using three-dimensional quantitative phase imaging and deep learning. We achieved 95% accuracy in classifying 19 bacterial species by exploiting the rich information in three-dimensional refractive index tomograms with a convolutional neural network classifier. Extensive analysis of the features extracted by the trained classifier was carried out, which supported that our classifier is capable of learning species-dependent characteristics. We also confirmed that utilizing three-dimensional refractive index tomograms was crucial for identification ability compared to two-dimensional imaging. This method, which does not require time-consuming culture, shows high feasibility for diagnosing patients with infectious diseases who would benefit from immediate and adequate antibiotic treatment.

Three-dimensional label-free observation of individual bacteria upon antibiotic treatment using optical diffraction tomography

“…This can also be used effectively to study the cell cycle of bacteria [52]. Because the presented method is general, it can be used for various applications of microbiology, ranging from bacteria classifications to cell dynamics in the bacterial infection scenarios [53,54].…”

Quantitative scoring of epithelial and mesenchymal qualities of cancer cells using machine learning and quantitative phase imaging

“…In single-cell studies, phase images of cells of interest could guide laser-capture microdissection to link observed behavior, morphology, and gene expression at a single cell level. Indeed, advanced machine learning techniques, including deep learning, 29 have recently been applied to isolate cell subpopulations based on unique phase features 6 and other phenotypic differences, 66,37 including metastatic versus primary cancer 67 and different types of nonactivated lymphocytes. 68 The phase/morphology score concept described here could be applied to support decision-making in intelligent cell sorting systems, such as flow cytometry with QPI, 69,34 to partition cells from a heterogeneous population into distinct morphological groups.…”

“…27 The use of parameters from QPI has recently been explored to assess shifts in population distributions of cell phase parameters, indicating altered phenotype, or to differentiate multiple bacteria species based on their single-cell profiling capability. 28,29 The effects of cell seeding density, 4 exposure to anticancer drugs, 30,31 and other influences on cell phenotype [32][33][34] have been robustly evaluated with QPI. Quantitative imaging and machine learning have the potential to save time, labor, and reduce human error in phenotypic profiling, which could help pathologists and scientists to accurately detect circulating tumor cells, 35 classify cancer cells, 36,37 evaluate the metastatic potential of cancer cells, 38 and assess cancer drug resistance.…”

Computational cytometer based on magnetically modulated coherent imaging and deep learning

“…The recent advances in machine learning and, specifically, deep learning have pushed the frontiers of biomedical imaging and image analysis 25–38 , enabling rapid and accurate pathogen detection 39–42 and computer-assisted diagnostic methods 43–47 . Powered by deep learning, we demonstrate here that speckle imaging using lensless chip-scale microscopy can be employed for the specific and sensitive detection of rare cells in blood with low cost and high throughput.…”