Label‐free surface‐enhanced Raman spectroscopy coupled with machine learning algorithms in pathogenic microbial identification: Current trends, challenges, and perspectives

Deleted Journal, Год журнала: 2024, Номер 2(3)

Опубликована: Март 5, 2024

Abstract Infectious diseases caused by microbial pathogens remain a primary contributor to global health burdens. Prompt control and effective prevention of these are critical for public medical diagnostics. Conventional detection methods suffer from high complexity, low sensitivity, poor selectivity. Therefore, developing rapid reliable pathogen has become imperative. Surface‐enhanced Raman Spectroscopy (SERS), as an innovative non‐invasive diagnostic technique, holds significant promise in pathogenic microorganism due its rapid, reliable, cost‐effective advantages. This review comprehensively outlines the fundamental theories (RS) with focus on label‐free SERS strategy, reporting latest advancements technique detecting bacteria, viruses, fungi clinical settings. Furthermore, we emphasize application machine learning algorithms spectral analysis. Finally, challenges faced probed, prospective development is discussed.

Язык: Английский

---

Machine Learning-Assisted Raman Spectroscopy and SERS for Bacterial Pathogen Detection: Clinical, Food Safety, and Environmental Applications

Chemosensors, Год журнала: 2024, Номер 12(7), С. 140 - 140

Опубликована: Июль 15, 2024

Detecting pathogenic bacteria and their phenotypes including microbial resistance is crucial for preventing infection, ensuring food safety, promoting environmental protection. Raman spectroscopy offers rapid, seamless, label-free identification, rendering it superior to gold-standard detection techniques such as culture-based assays polymerase chain reactions. However, its practical adoption hindered by issues related weak signals, complex spectra, limited datasets, a lack of adaptability characterization bacterial pathogens. This review focuses on addressing these with recent breakthroughs enabled machine learning (ML), particularly deep methods. Given the regulatory requirements, consumer demand safe products, growing awareness risks pathogens, this study emphasizes pathogen in clinical, settings. Here, we highlight use convolutional neural networks analyzing clinical data surface enhanced sensitizing early rapid pathogens safety potential risks. Deep methods can tackle adequate datasets across diverse samples. We pending future research directions needed accelerating real-world impacts ML-enabled diagnostics accurate diagnosis surveillance critical fields.

Язык: Английский

---

Key steps for improving bacterial SERS signals in complex samples: Separation, recognition, detection, and analysis

Talanta, Год журнала: 2023, Номер 268, С. 125281 - 125281

Опубликована: Окт. 7, 2023

Язык: Английский

---

Label-Free Identification of Spore-Forming Bacteria Using Ultrabroadband Multiplex Coherent Anti-Stokes Raman Scattering Microspectroscopy

The Journal of Physical Chemistry B, Год журнала: 2023, Номер 127(9), С. 1940 - 1946

Опубликована: Фев. 23, 2023

Spore-forming bacteria accumulate dipicolinic acid (DPA) to form spores survive in extreme environments. Vibrational spectroscopy is widely used detect DPA and elucidate the existence of bacteria, while vegetative cells, another spore-forming have not been studied extensively. Herein, we applied coherent anti-Stokes Raman scattering (CARS) microscopy spectroscopically identify both cells without staining or molecular tagging. The were identified by strong CARS signals due DPA. Furthermore, observed bright spots image at 1735 cm–1. contained species with C=O bonds because this vibrational mode was associated carbonyl group. One candidate diketopimelic (DKP), a precursor. This hypothesis verified comparing spectrum obtained that DKP analogue (ketopimelic acid) result DFT calculation. results indicate cell sporulation process. spectra can be monitor maturation preformation spores.

Язык: Английский

---

Artificial intelligence-aided rapid and accurate identification of clinical fungal infections by single-cell Raman spectroscopy

Frontiers in Microbiology, Год журнала: 2023, Номер 14

Опубликована: Март 22, 2023

Integrating artificial intelligence and new diagnostic platforms into routine clinical microbiology laboratory procedures has grown increasingly intriguing, holding promises of reducing turnaround time cost maximizing efficiency. At least one billion people are suffering from fungal infections, leading to over 1.6 million mortality every year. Despite the increasing demand for diagnosis, current approaches suffer manual bias, long cultivation (from days months), low sensitivity (only 50% produce positive cultures). Delayed inaccurate treatments consequently lead higher hospital costs, mobility rates. Here, we developed single-cell Raman spectroscopy achieve rapid identification infectious fungi. The classification between fungi bacteria infections was initially achieved with 100% specificity using spectra (SCRS). Then, constructed a dataset isolates obtained 94 patients, consisting 115,129 SCRS. By training model an optimized feedback loop, just 5 cells per patient (acquisition 2 s cell) made most accurate classification. This protocol accuracies at species level. transformed assessing samples urinary tract infection, obtaining correct diagnosis raw sample-to-result within 1 h.

Язык: Английский

---

Prospects of single-cell nuclear magnetic resonance spectroscopy with quantum sensors

Current Opinion in Biotechnology, Год журнала: 2023, Номер 83, С. 102975 - 102975

Опубликована: Авг. 11, 2023

Single-cell analysis can unravel functional heterogeneity within cell populations otherwise obscured by ensemble measurements. However, noninvasive techniques that probe chemical entities and their dynamics are still lacking. This challenge could be overcome novel sensors based on nitrogen-vacancy (NV) centers in diamond, which enable nuclear magnetic resonance (NMR) spectroscopy unprecedented sample volumes. In this perspective, we briefly introduce NV-based quantum sensing review the progress made microscale NV-NMR spectroscopy. Last, discuss approaches to enhance sensitivity of NV magnetometers detect biologically relevant concentrations provide a roadmap toward application single-cell analysis.

Язык: Английский

---

Rapid identification and drug resistance screening of respiratory pathogens based on single-cell Raman spectroscopy

Frontiers in Microbiology, Год журнала: 2023, Номер 14

Опубликована: Янв. 26, 2023

Respiratory infections rank fourth in the global economic burden of disease. Lower respiratory tract are leading cause death low-income countries. The rapid identification pathogens causing lower to help guide use antibiotics can reduce mortality patients with infections. Single-cell Raman spectroscopy is a “whole biological fingerprint” technique that be used identify microbial samples. It has advantages no marking and fast non-destructive testing. In this study, single-cell was collect spectral data six pathogen isolates. T-distributed stochastic neighbor embedding (t-SNE) isolation analysis algorithm compare differences between pathogens. eXtreme Gradient Boosting (XGBoost) establish phenotype database model. classification accuracy isolated samples 93–100%, clinical more than 80%. Combined heavy water labeling technology, drug resistance determined. study showed spectroscopy–D 2 O (SCRS–D O) could rapidly within h.

Язык: Английский

---

Unveiling the efficacy of a bulk Raman spectra-based model in predicting single cell Raman spectra of microorganisms

Heliyon, Год журнала: 2024, Номер 10(6), С. e27824 - e27824

Опубликована: Март 1, 2024

In a previous publication, we trained predictive models based on Raman bulk spectra of microorganisms placed silicon dioxide protected silver mirror slide to make predictions for new spectra, unknown the models, different substrate, namely stainless steel. Now have combined large sections this data and convolutional neural network (CNN) single cell spectra. We show that database microbial material is conditionally suited same species in terms cells. Data 13 (bacteria yeasts) were used. Two could be identified 90% correctly five other 71%–88%. The six remaining predicted by only 0%–49%. Especially stronger fluorescence compared cells but also photodegradation carotenoids are some effects can complicate data. results helpful assessing universal tools or databases.

Язык: Английский

---

Improving Machine Learning–Based Bacterial Discrimination by Learning Single‐Cell Raman Data From Multiple Growth Phases

Journal of Raman Spectroscopy, Год журнала: 2025, Номер unknown

Опубликована: Март 22, 2025

ABSTRACT Bacterial discrimination using single‐cell Raman spectroscopy and machine/deep learning techniques has been widely explored for promising applications in medical, environmental, food sciences. To construct a machine‐learning model that can achieve highly accurate robust of bacteria real‐world samples, data consisting spectra bacterial cells acquired under various physiological conditions are essential. Despite much effort to study the effects growth phase on discrimination, it is not yet fully elucidated which phase(s) needs be included training efficiently improve accuracy what phase‐dependent changes cellular components underlie discrimination. Here, we used random forest (RF), an ensemble machine method, discriminate six species, including both Gram‐positive Gram‐negative bacteria, at five different phases ranging from lag late stationary phases. We compared four RF classification models were trained one (either midexponential or stationary), two (midexponential all The species built distinctly exceeded 80% with marked increase 24% 32.5% relative single phase. This was greater than found going (13%). also revealed bands relatively invariant (e.g., proteins) specific DNA/RNA intracellular storage materials) important attaining present provides simple effective way good performance, could extended other such as nutrient, temperature, pH.

Язык: Английский

---

Machine Learning Approaches in Label-Free Small Extracellular Vesicles Analysis with Surface-Enhanced Raman Scattering (SERS) for Cancer Diagnostics

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Фев. 22, 2024

Abstract Early diagnosis remains of pivotal importance in reducing patient morbidity and mortality cancer. To this end, liquid biopsy is emerging as a tool to perform broad cancer screenings. Small extracellular vesicles (sEVs), also called exosomes, found bodily fluids can serve important biomarkers these Our group has recently developed label-free electrokinetic microchip purify sEVs from blood. Herein, we demonstrate the feasibility integrate approach with surface-enhanced Raman scattering (SERS) analysis. SERS be used characterized extracted through their vibrational fingerprint that changes depending on origin sEVs. While are not easily identified spectra, they modeled machine learning (ML) approaches. Common ML approaches field spectral analysis use dimensionality reduction method often function black box. avoid pitfall, Shapley additive explanations (SHAP) type explainable AI (XAI) bridges models human comprehension by calculating specific contribution individual features model’s predictions, directly correlating model/decisions original data. Using demonstrated proof-of-concept model predictive isolated sEVs, integrating device SERS. This work explores diagnostic complex data clinical samples, while reporting interpretable biochemical information.

Язык: Английский

---