Phenotypic Antimicrobial Susceptibility Testing with Deep Learning Video Microscopy

Abstract

Timely determination of antimicrobial susceptibility for a bacterial infection enables precision prescription, shortens treatment time, and helps minimize the spread of antibiotic resistant infections. Current antimicrobial susceptibility testing (AST) methods often take several days and thus impede these clinical and health benefits. Here, we present an AST method by imaging freely moving bacterial cells in urine in real time and analyzing the videos with a deep learning algorithm. The deep learning algorithm determines if an antibiotic inhibits a bacterial cell by learning multiple phenotypic features of the cell without the need for defining and quantifying each feature. We apply the method to urinary tract infection, a common infection that affects millions of people, to determine the minimum inhibitory concentration of pathogens from both bacteria spiked urine and clinical infected urine samples for different antibiotics within 30 min and validate the results with the gold standard broth macrodilution method. The deep learning video microscopy-based AST holds great potential to contribute to the solution of increasing drug-resistant infections.

Reagent-Free and Rapid Assessment of T Cell Activation State Using Diffraction Phase Microscopy and Deep Learning

“…35 The deep learning model is demonstrated to discern uninhibited bacterial cells from those cells inhibited by an antibiotic without necessitating conventional image processing. 35 A convolution neural network (CNN) type deep learning formalism was developed and trained for this data set. By testing various different network features, a 15-layer, comprised of 1 image input layer, 3 convolution 2D layers, 3 batch normalization, 3 ReLU, 2 2D max pooling, 1 fully connected, 1 softmax, and 1 classification layer.…”

Very rapid flow cytometric assessment of antimicrobial susceptibility during the apparent lag phase of microbial (re)growth

“…This notion probably stems from the fact that changes in OD observable by the naked eye in laboratory cultures [77] are indeed quite sluggish. However, the very few papers that have studied this in any detail [63, 84,85,95,97,98,102,103] have found that changes in expression profiles (albeit mainly measured at a bulk level) actually occur on a very rapid timescale indeed, possibly in 4 minutes or less following reinoculation into a rich growth medium. For antibiotics to have an observable, and in terms of sensitivity to them a differentially observable, effect on cells, the cells need to be in a replicative state.…”

“…The timescale in the plots of Madar and colleagues [95] does not admit quite such precise deconvolution, but responses in M9 with casamino acids (referred to as 'immediate') are consistent with a period of less than 10 min. Hong and colleagues recently detected such changes in under 30 min using stimulated Raman imaging [101], Yu and colleagues could do so with video microscopy [102], and Schoepp et al [103] used molecular detection of suitable transcripts. In view of the above, and recognising that bacteria in UTIs may actually be growing (albeit slowly) and not in a 'stationary' phase, we decided to assess the ability of quantitative flow cytometry to determine bacterial cell numbers, and the effects of antibiotics thereon, on as rapid a timescale as possible.…”

Cantilever Sensors for Rapid Optical Antimicrobial Sensitivity Testing

“…This has been achieved by miniaturizing the volume observed using microfluidics, [9][10][11] measuring mass or mechanical changes, [11][12][13] or by exploiting machine learning techniques for video tracking analysis of single cells. [14][15][16] Despite advances in the detection limit, and speed of testing, these are mostly complex set-ups, which remain far from point of care.…”