Dataset Information

Deep Learning-Based Culture-Free Bacteria Detection in Urine Using Large-Volume Microscopy.

ABSTRACT: Bacterial infections, increasingly resistant to common antibiotics, pose a global health challenge. Traditional diagnostics often depend on slow cell culturing, leading to empirical treatments that accelerate antibiotic resistance. We present a novel large-volume microscopy (LVM) system for rapid, point-of-care bacterial detection. This system, using low magnification (1-2×), visualizes sufficient sample volumes, eliminating the need for culture-based enrichment. Employing deep neural networks, our model demonstrates superior accuracy in detecting uropathogenic Escherichia coli compared to traditional machine learning methods. Future endeavors will focus on enriching our datasets with mixed samples and a broader spectrum of uropathogens, aiming to extend the applicability of our model to clinical samples.

SUBMITTER: Iriya R

PROVIDER: S-EPMC10887190 | biostudies-literature | 2024 Feb

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Deep Learning-Based Culture-Free Bacteria Detection in Urine Using Large-Volume Microscopy.

Iriya Rafael R Braswell Brandyn B Mo Manni M Zhang Fenni F Haydel Shelley E SE Wang Shaopeng S

Biosensors 20240205 2

Bacterial infections, increasingly resistant to common antibiotics, pose a global health challenge. Traditional diagnostics often depend on slow cell culturing, leading to empirical treatments that accelerate antibiotic resistance. We present a novel large-volume microscopy (LVM) system for rapid, point-of-care bacterial detection. This system, using low magnification (1-2×), visualizes sufficient sample volumes, eliminating the need for culture-based enrichment. Employing deep neural networks, ...[more]

PMID: 38392008

Similar Datasets

Project description:Urine cell-free DNA (cfDNA) is a valuable non-invasive biomarker with broad potential clinical applications, but there is no consensus on its optimal pre-analytical methodology, including the DNA extraction step. Due to its short length (majority of fragments <100 bp) and low concentration (ng/mL), urine cfDNA is not efficiently recovered by conventional silica-based extraction methods. To maximize sensitivity of urine cfDNA assays, we developed an ultrasensitive hybridization method that uses sequence-specific oligonucleotide capture probes immobilized on magnetic beads to improve extraction of short cfDNA from large-volume urine samples. Our hybridization method recovers near 100% (95% CI: 82.6-117.6%) of target-specific DNA from 10 mL urine, independent of fragment length (25-150 bp), and has a limit of detection of ≤5 copies of double-stranded DNA (0.5 copies/mL). Pairing hybridization with an ultrashort qPCR design, we can efficiently capture and amplify fragments as short as 25 bp. Our method enables amplification of cfDNA from 10 mL urine in a single qPCR well, tolerates variation in sample composition, and effectively removes non-target DNA. Our hybridization protocol improves upon both existing silica-based urine cfDNA extraction methods and previous hybridization-based sample preparation protocols. Two key innovations contribute to the strong performance of our method: a two-probe system enabling recovery of both strands of double-stranded DNA and dual biotinylated capture probes, which ensure consistent, high recovery by facilitating optimal probe density on the bead surface, improving thermostability of the probe-bead linkage, and eliminating interference by endogenous biotin. We originally designed the hybridization method for tuberculosis diagnosis from urine cfDNA, but expect that it will be versatile across urine cfDNA targets, and may be useful for other cfDNA sample types and applications beyond cfDNA. To make our hybridization method accessible to new users, we present a detailed protocol and straightforward guidelines for designing new capture probes.

Project description:Detection and identification of pathogenic bacteria isolated from biological samples (blood, urine, sputum, etc.) are crucial steps in accelerated clinical diagnosis. However, accurate and rapid identification remain difficult to achieve due to the challenge of having to analyse complex and large samples. Current solutions (mass spectrometry, automated biochemical testing, etc.) propose a trade-off between time and accuracy, achieving satisfactory results at the expense of time-consuming processes, which can also be intrusive, destructive and costly. Moreover, those techniques tend to require an overnight subculture on solid agar medium delaying bacteria identification by 12-48 hours, thus preventing rapid prescription of appropriate treatment as it hinders antibiotic susceptibility testing. In this study, lens-free imaging is presented as a possible solution to achieve a quick and accurate wide range, non-destructive, label-free pathogenic bacteria detection and identification in real-time using micro colonies (10-500 μm) kinetic growth pattern combined with a two-stage deep learning architecture. Bacterial colonies growth time-lapses were acquired thanks to a live-cell lens-free imaging system and a thin-layer agar media made of 20 μl BHI (Brain Heart Infusion) to train our deep learning networks. Our architecture proposal achieved interesting results on a dataset constituted of seven different pathogenic bacteria-Staphylococcus aureus (S. aureus), Enterococcus faecium (E. faecium), Enterococcus faecalis (E. faecalis), Staphylococcus epidermidis (S. epidermidis), Streptococcus pneumoniae R6 (S. pneumoniae), Streptococcus pyogenes (S. pyogenes), Lactococcus Lactis (L. Lactis). At T = 8h, our detection network reached an average 96.0% detection rate while our classification network precision and sensitivity averaged around 93.1% and 94.0% respectively, both were tested on 1908 colonies. Our classification network even obtained a perfect score for E. faecalis (60 colonies) and very high score for S. epidermidis at 99.7% (647 colonies). Our method achieved those results thanks to a novel technique coupling convolutional and recurrent neural networks together to extract spatio-temporal patterns from unreconstructed lens-free microscopy time-lapses.

Dataset Information

Deep Learning-Based Culture-Free Bacteria Detection in Urine Using Large-Volume Microscopy.

Publications

Deep Learning-Based Culture-Free Bacteria Detection in Urine Using Large-Volume Microscopy.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets