Antimicrobial susceptibility testing with large-volume light scattering imaging and deep learning video microscopy
A method for deep learning video microscopy-based antimicrobial susceptibility testing of a bacterial strain in a sample by acquiring image sequences of individual bacterial cells of the bacterial strain in a subject sample before, during, and after exposure to each antibiotic at different concentrations. The image sequences are compressed into static images while preserving essential phenotypic features. Data representing the static images are input into a pre-trained deep learning (DL) model which generates output data; and antimicrobial susceptibility for the bacterial strain is determined from the output data.
1 . A method for deep learning video microscopy-based antimicrobial susceptibility testing (DLVM-AST) of a bacterial strain in a sample including non-immobilized bacterial cells and an antibiotic, the method comprising:
acquiring an image sequence of individual bacterial cells of the bacterial strain in a subject sample with a phase contrast microscope before, during, and after exposure to each antibiotic at different concentrations, wherein the acquiring of the image sequence comprises capturing one or more phenotypic features including a motion feature of the non-immobilized bacterial cells in the image sequence, wherein superposition of multiple images of the image sequence yields a trace representing the motion feature;
compressing the image sequence into one or more trace images static images while preserving essential phenotypic features, wherein the one or more trace images are generated by superposition of multiple images of the image sequence, wherein the one or more trace images represent the motion feature;
inputting data representing the static images into a pre-trained deep learning (DL) model which generates output data; and
determining antimicrobial susceptibility for the bacterial strain from the output data.
2 . The method of claim 1 further comprising obtaining a minimum inhibitory concentration (MIC) value for the bacterial strain.
3 . The method of claim 1 wherein the subject sample is a urine sample from a human patient.
4 . The method of claim 1 wherein said capturing of one or more phenotypic features further comprises capturing one or more of a division feature and a morphological feature of the non-immobilized bacterial cells in the image sequence.
5 . The method of claim 1 wherein the antibiotic is selected from the group consisting of Polymyxin B, streptomycin, ciprofloxacin, aztreonam, ampicillin, and combinations thereof.