Separation-free bacterial identification in arbitrary media via deep neural network-based SERS analysis

Broad-spectrum pathogenic bacteria SERS sensing with face-centered high-index facets Au CPNCs & microarray chips: A novel platform able to achieve dual-readout detection

Non-specificity and inadequate quantitative capability are the primary challenges faced by the surface-enhanced Raman scattering (SERS) technique, especially when it comes to detecting bacteria in real samples. Herein, a novel face-centered Au Convex Polyhedral Nanocrystal (Au CPNC) with high-index facets and its assembly Au CPNCs microarray chip were designed and fabricated to address these challenges, within the process where 4-mercaptophenylboronic acid (4-MPBA) was utilized as a multifunctional element. The as-prepared Au CPNC possesses anisotropic raised edges enjoying tunable localized surface plasmon resonance modes for SERS enhancement. Then we obtained long-region ordered Au CPNCs microarrays equipping even greater "hot spots" with a SERS enhancement factor (EF) up to 5.38 × 107. The constructed SERS probes excellently leveraged the outstanding SERS performance of Au CPNC and the superior functions of 4-MPBA, which enabled the differences among the bacterial "fingerprints" to be highlighted. Through partial least squares discriminant analysis (PLS-DA), we successfully identified Escherichia coli, Staphylococcus aureus, Pseudomonas aeruginosa and Listeria monocytogenes with achieving limits of detection (LODs) in spiked whole blood samples of 3, 1, 2, and 2 cfu/mL, respectively. Notably, the LODs for all samples by SERS mapping visual readout mode were as low as 10 cfu/mL. In practical applications, our method demonstrated its efficacy by 100 % accurately classifying (20 cases) of real blood samples. Altogether, the theoretical significance and application value of this study reside in providing fundamental insights and approaches for the development of pathogenic bacteria detection field.

Artificial intelligence in bacterial diagnostics and antimicrobial susceptibility testing: Current advances and future prospects

Recently, artificial intelligence (AI) has emerged as a transformative tool, enhancing the speed, accuracy, and scalability of bacterial diagnostics. This review explores the role of AI in revolutionizing bacterial detection and antimicrobial susceptibility testing (AST) by leveraging machine learning models, including Random Forest, Support Vector Machines (SVM), and deep learning architectures such as Convolutional Neural Networks (CNNs) and transformers. The integration of AI into these methods promises to address the current limitations of traditional techniques, offering a path toward more efficient, accessible, and reliable diagnostic solutions. In particular, AI-based approaches have demonstrated significant potential in resource-limited settings by enabling cost-effective and portable diagnostic solutions, reducing dependency on specialized infrastructure, and facilitating remote bacterial detection through smartphone-integrated platforms and telemedicine applications. This review highlights AI's transformative role in automating data analysis, minimizing human error, and delivering real-time diagnostic results, ultimately improving patient outcomes and optimizing healthcare efficiency. In addition, we not only examine the current advances in machine learning and deep learning but also review their applications in plate counting, mass spectrometry, morphology-based and motion-based microscopic detection, holographic microscopy, colorimetric and fluorescence detection, electrochemical sensors, Raman and Surface-Enhanced Raman Spectroscopy (SERS), and Atomic Force Microscopy (AFM) for bacterial diagnostics and AST. Finally, we discuss the future directions and potential advancements in AI-driven bacterial diagnostics.

Deep learning enabled open-set bacteria recognition using surface-enhanced Raman spectroscopy

Accurate bacterial identification is vital in medical and healthcare settings. Traditional methods, though reliable, are often time-consuming, underscoring the need for faster, more efficient alternatives. Deep learning-assisted Surface-enhanced Raman spectroscopy (SERS) offers a rapid and sensitive method, demonstrating high accuracy in bacterial identification. However, current deep learning models for bacterial SERS spectra classification typically operate under a closed-set paradigm, limiting their effectiveness when encountering bacterial species outside the training set. In response to this challenge, we propose a transformer-based neural network for open-set bacterial recognition using SERS spectra. Our model utilizes a combination of classification and reconstruction tasks, rejecting unknown species by analyzing reconstruction errors. Experimental results show that the proposed model outperforms traditional open-set recognition approaches, providing superior accuracy in both classifying known species and rejecting unknown ones. This study addresses the limitations of existing closed-set methods, improving the robustness of bacterial identification in real-world scenarios and demonstrating the potential of integrating SERS with transformer models for medical and healthcare applications.

Comparison of Raman spectroscopy with mass spectrometry for sequence typing of Acinetobacter baumannii strains: a single-center study

The rapid sequence typing (ST) of bacterial strains is crucial for effective nosocomial infection control and mitigating the spread of nosocomial pathogens, e.g., Acinetobacter baumannii. While accurate in identifying A. baumannii strains, current typing methods are often impractical in clinical settings due to their time-consuming nature. This study developed a novel approach, combining surface-enhanced Raman spectroscopy (SERS) with machine-learning (ML) algorithms, to construct predictive models for A. baumannii sequence typing based on SERS spectra. The objective was to assess the feasibility of this integrated method for efficient sequence typing of A. baumannii strains. Clinically isolated A. baumannii strains (N = 267) were collected from a single hospital between 2013 and 2023. Based on multilocus sequence typing, 39 STs of A. baumannii were identified. Then, a SERS spectral database for all these strains was constructed, and predictive models based on eight ML algorithms were developed to predict SERS signals to determine their STs, among which the support vector machine (SVM) model had the best performance (fivefold cross-validation = 99.74%). The typing capacity of the SERS-SVM method was compared with that of matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) for A. baumannii sequence typing, confirming the superiority of SERS-SVM over MALDI-TOF mass spectrometer. This pilot study lays the groundwork for employing the SERS-ML method to rapidly identify A. baumannii strain types in clinical laboratories, aiding in controlling bacterial pathogen transmission. Further studies are warranted to evaluate its potential in nosocomial surveillance systems, especially for rapidly identifying outbreaks within hospitals.

IMPORTANCE

The rapid and accurate sequence typing (ST) of bacterial pathogens is pivotal in controlling transmission within healthcare settings. Acinetobacter baumannii infection, known for its high transmissibility and drug resistance, presents a major challenge in nosocomial infection control. In this study, surface-enhanced Raman spectroscopy (SERS) was used to differentiate A. baumannii strains with distinct STs based on unique Raman spectral profiles. We then constructed and compared eight machine-learning models on SERS spectra to quickly identify bacterial STs. The results showed that the support vector machine model outperformed matrix-assisted laser desorption/ionization time-of-flight mass spectrometer in determining A. baumannii STs. This approach enables rapid identification of A. baumannii variants with different STs, supporting the early detection and control of nosocomial infections by this multidrug-resistant pathogen.

Identification of SERS Fingerprints for Diagnosis and Staging of Liver Injury

Acute liver injury is a common hepatic condition that, without timely diagnosis and treatment, can easily progress to life-threatening liver failure. Identifying key biomolecular changes in the pathogenesis of liver injury is critical for the precise diagnosis of hepatic injury. In this study, we identified novel molecular fingerprints associated with hepatic injury using artificial intelligence-assisted Surface-enhanced Raman spectroscopy (SERS) technology. By intravenous injection of PEG-modified Au nanoparticles, which efficiently accumulate in the liver, these nanoparticles act as in vivo computed tomography contrast agents, specifically delineating the three-dimensional structure of normal and injured liver tissues. Additionally, as a Raman-enhanced substrate, gold nanoparticles allow for the analysis of SERS signals in liver tissues enriched with these nanoparticles. Leveraging artificial intelligence technology, we achieved a classification accuracy of 96.45% for different degrees of liver injury. Molecular spectral analysis revealed that the ratio of the 1437-1000 cm-1 signal could serve as an SERS fingerprint correlated with hepatic injury. The integrated information system provides a valuable tool for the precise diagnosis of liver diseases.

A multilevel cooperative attention network of precise quantitative analysis for predicting ractopamine concentration via adaptive weighted feature selection and multichannel feature fusion

Surface enhanced Raman spectroscopy (SERS) holds great potential due to its rapid detection and high sensitivity. However, issues such as signal noise, fluctuations, and spectral shifts can negatively impact its performance in detecting ractopamine in pork. Hierarchical Gradient Aware Spectral Network (HGASNet) was proposed to address these issues. The key innovations are the Spectral Gradient Weighted Attention (SGWA) and Multi-Channel Peak Attention Mechanism (MCPAM) modules within HGASNet. The SGWA module dynamically adjusts feature weights, enhancing sensitivity to critical Raman spectral shifts. Meanwhile, MCPAM leverages multi-channel attention to better capture long range dependencies and fuse global and local information. Additionally, HGASNet's hierarchical structure incrementally extracts and integrates features at various levels, enabling the model to focus on global features in the higher layers while preserving fine-grained local details in the lower layers. Experimental results show HGASNet outperforming existing approaches, achieving R2 of 0.9972, RMSRE of 0.0413, RMSLE of 0.0548, sMAPE of 6.47 %, and MAPE of 6.72 %.

In situ surface-enhanced Raman spectroscopy for membrane protein analysis and sensing

Membrane proteins are involved in a variety of dynamic cellular processes and exploration of the structural basis of membrane proteins is of significance for a better understanding of their functions. In situ analysis of membrane proteins and their dynamics is, however, challenging for conventional techniques. Surface-enhanced Raman spectroscopy (SERS) is powerful in protein structural characterization, allowing for sensitive, in-situ and real-time identification and dynamic monitoring under physiological conditions. In this review, the applications of SERS in probing membrane proteins are outlined, discussed and prospected. It starts with a brief introduction to membrane proteins, SERS theories and SERS-based strategies that commonly-used for membrane proteins. How to assemble phospholipid biolayers on SERS-active materials is highlighted, followed by respectively discussing about direct and indirect strategies for membrane protein sensing. SERS-based monitoring of protein-ligand interactions is finally introduced and its potential in biomedical applications is discussed in detail. The review ends with critical discussion about current challenges and limitations of this research field, and the promising perspectives in both fundamental and applied sciences.

Carbon nanomaterial-based aptasensors for rapid detection of foodborne pathogenic bacteria

Each year, millions of people suffer from foodborne illness due to the consumption of food contaminated with pathogenic bacteria, which severely challenges global health. Therefore, it is essential to recognize foodborne pathogens swiftly and correctly. However, conventional detection techniques for bacterial pathogens are labor-intensive, low selectivity, and time-consuming, highlighting a notable knowledge gap. A novel approach, aptamer-based biosensors (aptasensors) linked to carbon nanomaterials (CNs), has shown the potential to overcome these limitations and provide a more reliable method for detecting bacterial pathogens. Aptamers, short single-stranded DNA (ssDNA)/RNA molecules, serve as bio-recognition elements (BRE) due to their exceptionally high affinity and specificity in identifying foodborne pathogens such as Salmonella spp., Escherichia coli (E. coli), Listeria monocytogenes, Campylobacter jejuni, and other relevant pathogens commonly associated with foodborne illnesses. Carbon nanomaterials' high surface area-to-volume ratio contributes unique characteristics crucial for bacterial sensing, as it improves the binding capacity and signal amplification in the design of aptasensors. Furthermore, aptamers can bind to CNs and create aptasensors with improved signal specificity and sensitivity. Hence, this review intends to critically review the current literature on developing aptamer functionalized CN-based biosensors by transducer optical and electrochemical for detecting foodborne pathogens and explore the advantages and challenges associated with these biosensors. Aptasensors conjugated with CNs offers an efficient tool for identifying foodborne pathogenic bacteria that is both precise and sensitive to potentially replacing complex current techniques that are time-consuming.

Integrating machine learning and biosensors in microfluidic devices: A review

Microfluidic devices are increasingly widespread in the literature, being applied to numerous exciting applications, from chemical research to Point-of-Care devices, passing through drug development and clinical scenarios. Setting up these microenvironments, however, introduces the necessity of locally controlling the variables involved in the phenomena under investigation. For this reason, the literature has deeply explored the possibility of introducing sensing elements to investigate the physical quantities and the biochemical concentration inside microfluidic devices. Biosensors, particularly, are well known for their high accuracy, selectivity, and responsiveness. However, their signals could be challenging to interpret and must be carefully analysed to carry out the correct information. In addition, proper data analysis has been demonstrated even to increase biosensors' mentioned qualities. To this regard, machine learning algorithms are undoubtedly among the most suitable approaches to undertake this job, automatically learning from data and highlighting biosensor signals' characteristics at best. Interestingly, it was also demonstrated to benefit microfluidic devices themselves, in a new paradigm that the literature is starting to name "intelligent microfluidics", ideally closing this benefic interaction among these disciplines. This review aims to demonstrate the advantages of the triad paradigm microfluidics-biosensors-machine learning, which is still little used but has a great perspective. After briefly describing the single entities, the different sections will demonstrate the benefits of the dual interactions, highlighting the applications where the reviewed triad paradigm was employed.

Leveraging Broad-Spectrum Fluorescence Data and Machine Learning for High-Accuracy Bacterial Species Identification

Rapid and accurate identification of bacterial species is essential for the effective treatment of infectious diseases and suppression of antibiotic-resistant strains. The unique autofluorescence properties of bacterial cells are exploited for rapid and cost-effective identification that is suitable for point-of-care applications. Fluorescence spectroscopy is combined with machine learning to improve the diagnostic accuracy. Good training data for machine learning can be obtained to achieve the same diagnostic accuracy for bacterial species as when each wavelength is measured in detail over a broad spectral width. Experiments were performed testing 14 bacterial strains. The excitation-emission matrix was analyzed, and Bayesian optimization was used to identify the most effective combinations of wavelengths. The results showed that fluorescence spectra using three specific excitation light regions or excitation spectra using two broad fluorescence detection regions could be used as supervised data to realize diagnostic accuracy comparable to that obtained with more complex instruments.

Deep learning-assisted monitoring of trastuzumab efficacy in HER2-Overexpressing breast cancer via SERS immunoassays of tumor-derived urinary exosomal biomarkers

Monitoring drug efficacy is significant in the current concept of companion diagnostics in metastatic breast cancer. Trastuzumab, a drug targeting human epidermal growth factor receptor 2 (HER2), is an effective treatment for metastatic breast cancer. However, some patients develop resistance to this therapy; therefore, monitoring its efficacy is essential. Here, we describe a deep learning-assisted monitoring of trastuzumab efficacy based on a surface-enhanced Raman spectroscopy (SERS) immunoassay against HER2-overexpressing mouse urinary exosomes. Individual Raman reporters bearing the desired SERS tag and exosome capture substrate were prepared for the SERS immunoassay; SERS tag signals were collected to prepare deep learning training data. Using this deep learning algorithm, various complicated mixtures of SERS tags were successfully quantified and classified. Exosomal antigen levels of five types of cell-derived exosomes were determined using SERS-deep learning analysis and compared with those obtained via quantitative reverse transcription polymerase chain reaction and western blot analysis. Finally, drug efficacy was monitored via SERS-deep learning analysis using urinary exosomes from trastuzumab-treated mice. Use of this monitoring system should allow proactive responses to any treatment-resistant issues.

Recent Advances in Bacterial Detection Using Surface-Enhanced Raman Scattering

Rapid identification of microorganisms with a high sensitivity and selectivity is of great interest in many fields, primarily in clinical diagnosis, environmental monitoring, and the food industry. For over the past decades, a surface-enhanced Raman scattering (SERS)-based detection platform has been extensively used for bacterial detection, and the effort has been extended to clinical, environmental, and food samples. In contrast to other approaches, such as enzyme-linked immunosorbent assays and polymerase chain reaction, SERS exhibits outstanding advantages of rapid detection, being culture-free, low cost, high sensitivity, and lack of water interference. This review aims to cover the development of SERS-based methods for bacterial detection with an emphasis on the source of the signal, techniques used to improve the limit of detection and specificity, and the application of SERS in high-throughput settings and complex samples. The challenges and advancements with the implementation of artificial intelligence (AI) are also discussed.

Multiplex Detection of Foodborne Pathogens using 3D Nanostructure Swab and Deep Learning-Based Classification of Raman Spectra

Proactive management of foodborne illness requires routine surveillance of foodborne pathogens, which requires developing simple, rapid, and sensitive detection methods. Here, a strategy is presented that enables the detection of multiple foodborne bacteria using a 3D nanostructure swab and deep learning-based Raman signal classification. The nanostructure swab efficiently captures foodborne pathogens, and the portable Raman instrument directly collects the Raman signals of captured bacteria. a deep learning algorithm has been demonstrated, 1D convolutional neural network with binary labeling, achieves superior performance in classifying individual bacterial species. This methodology has been extended to mixed bacterial populations, maintaining accuracy close to 100%. In addition, the gradient-weighted class activation mapping method is used to provide an investigation of the Raman bands for foodborne pathogens. For practical application, blind tests are conducted on contaminated kitchen utensils and foods. The proposed technique is validated by the successful detection of bacterial species from the contaminated surfaces. The use of a 3D nanostructure swab, portable Raman device, and deep learning-based classification provides a powerful tool for rapid identification (≈5 min) of foodborne bacterial species. The detection strategy shows significant potential for reliable food safety monitoring, making a meaningful contribution to public health and the food industry.

Surface-functionalized SERS platform for deep learning-assisted diagnosis of Alzheimer's disease

Early diagnosis of Alzheimer's disease is crucial to stall the deterioration of brain function, but conventional diagnostic methods require complicated analytical procedures or inflict acute pain on the patient. Then, label-free Surface-enhanced Raman spectroscopy (SERS) analysis of blood-based biomarkers is a convenient alternative to rapidly obtain spectral information from biofluids. However, despite the rapid acquisition of spectral information from biofluids, it is challenging to distinguish spectral features of biomarkers due to interference from biofluidic components. Here, we introduce a deep learning-assisted, SERS-based platform for separate analysis of blood-based amyloid β (1-42) and metabolites, enabling the diagnosis of Alzheimer's disease. SERS substrates consisting of Au nanowire arrays are fabricated and functionalized in two characteristic ways to compare the validity of different Alzheimer's disease biomarkers measured on our SERS system. The 6E10 antibody is immobilized for the capture of amyloid β (1-42) and analysis of its oligomerization process, while various self-assembled monolayers are attached for different dipole interactions with blood-based metabolites. Ultimately, SERS spectra of blood plasma of Alzheimer's disease patients and human controls are measured on the substrates and classified via advanced deep learning techniques that automatically extract informative features to learn generalizable representations. Accuracies up to 99.5% are achieved for metabolite-based analyses, which are verified with an explainable artificial intelligence technique that identifies key spectral features used for classification and for deducing significant biomarkers.

A novel multifunctional SERS microfluidic sensor based on ZnO/Ag nanoflower arrays for label-free ultrasensitive detection of bacteria

This study proposes a microfluidic platform for rapid enrichment and ultrasensitive SERS detection of bacteria. The platform comprises ZnO nanoflower arrays decorated with silver nanoparticles to enhance the SERS sensitivity. The ZnO nanoflower array substrate with a 3D reticular columnar structure is prepared using the hydrothermal method. SEM analysis depicts the 3.05 μm gap distribution of the substrate array to intercept the most bacteria in the particle sizes range of 0.5 to 3 μm. Then, silver nanoparticles are deposited on the ZnO nano-array surface by liquid evaporation self-assembly. TEM and SEM analysis indicate nanosize of Ag particles, evenly distributed on the substrate, enhancing the SERS efficiency and improving sensing reproducibility. The probe molecules (R6G) are tested to demonstrate the high SERS activity of the proposed microfluidic sensor. Then, Escherichia coli, Staphylococcus aureus, Enterococcus faecalis, and Bacillus subtilis are selected, demonstrating the sensor's excellent bacterial capture and sensitive recognition capabilities, with a detection limit as low as 102 CFU mL-1. Additionally, the antibacterial properties of ZnO/Ag heterojunction nanostructures are studied, suggesting their ability to inactivate bacteria. Compared with the traditional Au-enhanced chip, the sensor preparation is easy, safe, reliable, and low-cost. Moreover, the ZnO nano-array exhibits a large specific surface area, high interception ability, stronger and uniform SERS performance, and effective and reliable detection of trace pathogens. This work provides potential future ZnO/Ag microfluidic SERS sensor applications for rapid, unlabeled, and trace pathogens detection in clinical and environmental applications, potentially achieving breakthroughs in early detection, prevention, and treatment.

Simultaneous detection of SARS-CoV-2 S1 protein by using flexible electrochemical and Raman enhancing biochip

Flexible laser-scribed graphene (LSG) substrates with gold nanoislands have been developed as biochips for in situ electrochemical (EC) and surface-enhanced Raman scattering (SERS) biodetection (biomolecules and viral proteins). A flexible biochip was fabricated using CO2 laser engraving polyimide (PI) films to form a 3D porous graphene-like nanostructure. Gold nanoislands were deposited on the LSG substrates to enhance the intensity of the Raman signals. Moreover, the addition of auxiliary and reference electrodes induced a dual-function EC-SERS biochip with significantly enhanced detection sensitivity. The biochip could selectively and easily capture SARS-CoV-2 S1 protein through the SARS-CoV-2 S1 antibody immobilized on EC-SERS substrates using 1-ethyl-(3-dimethylaminopropyl)carbodiimide (EDC) and N-hydroxysuccinimide (NHS). The grafted antibody specifically bound to SARS-CoV-2, resulting in a significant increase in the SERS signal of the target analyte. The limit of detection (LOD) of the SARS-CoV-2 S1 protein was 5 and 100 ng/mL by using EC and SERS detection, respectively. Although the LOD of the SARS-CoV-2 S1 protein detected using SERS is only 100 ng/mL, it can provide fingerprint information for identification. To improve the LOD, EC detection was integrated with SERS detection. The three-electrode detection chip enables the simultaneous detection of SERS and EC signals, which provides complementary information for target identification. The dual-functional detection technology demonstrated in this study has great potential for biomedical applications, such as the rapid and sensitive detection of SARS-CoV-2.

Qualitative and quantitative studies of phthalates in extra virgin olive oil (EVOO) by surface-enhanced Raman spectroscopy (SERS) combined with long short term memory (LSTM) neural network

Phthalates are commonly used plasticizers in the plastics industry, and have received extensive attention due to their reproductive toxicity. Since phthalates are lipophilic solutions, phthalates can easily migrate from packaging to edible oils. This study synthesized stable and sensitive Gold Nanostars as SERS substrates to conduct qualitative and quantitative analysis of two common phthalates, dibutyl phthalate and di(2-ethylhexyl) phthalate. Two ethanol standard solutions and actual oil solutions of phthalates at different concentrations (10, 5, 1, 0.5, 0.1, 0.02 mg/kg) were prepared. After dimension reduction, LSTM achieved the accuracy of 98% for pure EVOO and EVOO adulterated with different types of phthalates. In terms of quantification, LSTM demonstrates great predictive performance with Rp2 greater than 0.97 and the ratio of performance to deviation greater than 5. These results have certain guiding significance for the analysis of plasticizers in edible oil.

Machine learning enabled multiplex detection of periodontal pathogens by surface-enhanced Raman spectroscopy

Periodontitis is a chronic inflammation of the periodontium caused by a persistent bacterial infection, resulting in destruction of the supporting structures of teeth. Analysis of microbial composition in saliva can inform periodontal status. Actinobacillus actinomycetemcomitans (Aa), Porphyromonas gingivalis (Pg), and Streptococcus mutans (Sm) are among reported periodontal pathogens, and were used as model systems in this study. Our atomic force microscopic (AFM) study revealed that these pathogens are biological nanorods with dimensions of 0.6-1.1 μm in length and 500-700 nm in width. Current bacterial detection methods often involve complex preparation steps and require labeled reporting motifs. Employing surface-enhanced Raman spectroscopy (SERS), we revealed cell-type specific Raman signatures of these pathogens for label-free detection. It overcame the complexity associated with spectral overlaps among different bacterial species, relying on high signal-to-noise ratio (SNR) spectra carefully collected from pure species samples. To enable simple, rapid, and multiplexed detection, we harnessed advanced machine learning techniques to establish predictive models based on a large set of raw spectra of each bacterial species and their mixtures. Using these models, given a raw spectrum collected from a bacterial suspension, simultaneous identification of all three species in the test sample was achieved at 95.6 % accuracy. This sensing modality can be applied to multiplex detection of a broader range and a larger set of periodontal pathogens, paving the way for hassle-free detection of oral bacteria in saliva with little to no sample preparation.

Deep learning enhanced multiplex detection of viable foodborne pathogens in digital microfluidic chip

Culture plating is worldwide accepted as the gold standard for quantifying viable foodborne pathogens. However, it is time-consuming (1-2 days) and requires specialized laboratory and personnel. This study reported a deep learning enhanced digital microfluidic platform for multiplex detection of viable foodborne pathogens. The new method used a Time-Lapse images driven EfficientNet-Transformer Network (TLENTNet) to type and quantify the bacteria through spatiotemporal features of bacterial growth and digital enumeration of bacterial culture. First, the bacterial sample was prepared with LB medium and injected into a pre-vacuumed microfluidic chip with an array of 800 microwells to encapsulate at most one bacterium in each well. Then, a programmed sliding microscopic platform was used to scan all microwells every 15 min, capturing time-lapse images of bacterial growth within each microwell. Finally, the TLENTNet was used to facilitate bacterial typing and quantification. Under optimal conditions, this platform was able to detect four bacterial species (S.typhimurium, E. coli O157:H7, S. aureus and B. cereus) with an average accuracy of 97.72% and a detection limit of 63 CFU/mL in 7 h.

DBAN: An improved dual branch attention network combined with serum Raman spectroscopy for diagnosis of diabetic kidney disease

Diabetic kidney disease (DKD) is one of the most common kidney diseases worldwide. It is estimated that approximately 537 million adults worldwide have diabetes, and up to 30%-40% of diabetic patients are at risk of developing nephropathy. The pathogenesis of DKD is complex, and its onset is insidious. Currently, the clinical diagnosis of DKD primarily relies on the increase of urinary albumin and the decrease in glomerular filtration rate in diabetic patients. However, the excretion of urinary albumin is influenced by various factors, such as physical activity, infections, fever, and high blood glucose, making it challenging to achieve an objective and accurate diagnosis. Therefore, there is an urgent need to develop an efficient, fast, and low-cost auxiliary diagnostic technology for DKD. In this study, an improved Dual Branch Attention Network (DBAN) was developed to quickly identify DKD. Serum Raman spectroscopy samples were collected from 32 DKD patients and 32 healthy volunteers. The collected data were preprocessed using the adaptive iteratively reweighted penalized least squares (airPLS) algorithm, and the DBAN was used to classify the serum Raman spectroscopy data of DKD. The model consists of a dual branch structure that extracts features using Convolutional Neural Network (CNN) and bottleneck layer modules. The attention module allows the model to learn features specifically, and lateral connections are added between the dual branches to achieve multi-level and multi-scale fusion of shallow and deep features, as well as local and global features, improving the classification accuracy of the experiment. The results of the study showed that compared to traditional deep learning algorithms such as Artificial Neural Network (ANN), CNN, GoogleNet, ResNet, and AlexNet, our proposed DBAN classification model achieved 95.4% accuracy, 98.0% precision, 96.5% sensitivity, and 97.2% specificity, demonstrating the best classification performance. This is the best method for identifying DKD, and has important reference value for the diagnosis of DKD patients, as well as improving the accuracy of medical auxiliary diagnosis.

ZnO@Ag-Functionalized Paper-Based Microarray Chip for SERS Detection of Bacteria and Antibacterial and Photocatalytic Inactivation

Bacterial infections caused by pathogenic microorganisms have become a serious, widespread health concern. Thus, it is essential and required to develop a multifunctional platform that can rapidly and accurately determine bacteria and effectively inhibit or inactivate pathogens. Herein, a microarray SERS chip was successfully synthesized using novel metal/semiconductor composites (ZnO@Ag)-ZnO nanoflowers (ZnO NFs) decorated with Ag nanoparticles (Ag NPs) arrayed on a paper-based chip as a supporting substrate for in situ monitoring and photocatalytic inactivation of pathogenic bacteria. Typical Gram-positive Staphylococcus aureus and Gram-negative Escherichia coli and Vibrio parahemolyticus were selected as models. Partial least-squares discriminant analysis (PLS-DA) was performed to minimize the dimensionality of SERS spectra data sets and to develop a cost-effective identification model. The classification accuracy was 100, 97.2, and 100% for S. aureus, E. coli, and V. parahemolyticus, respectively. The antimicrobial activity of ZnO@Ag was proved by the microbroth dilution method, and the minimum inhibitory concentrations (MICs) of S. aureus, E. coli, and V. parahemolyticus were 40, 50, and 55 μg/mL, respectively. Meanwhile, it demonstrated remarkable photocatalytic performance under natural sunlight for the inactivation of pathogenic bacteria, and the inactivation rates for S. aureus, E. coli, and V. parahemolyticus were 100, 97.03 and 97.56%, respectively. As a result, the microarray chip not only detected the bacteria with high sensitivity but also confirmed the antibacterial and photocatalytic sterilization properties. Consequently, it offers highly prospective strategies for foodborne diseases caused by pathogenic bacteria.

Clinical screening of Nocardia in sputum smears based on neural networks

Objective

Nocardia is clinically rare but highly pathogenic in clinical practice. Due to the lack of Nocardia screening methods, Nocardia is often missed in diagnosis, leading to worsening the condition. Therefore, this paper proposes a Nocardia screening method based on neural networks, aiming at quick Nocardia detection in sputum specimens with low costs and thereby reducing the missed diagnosis rate.

Methods

Firstly, sputum specimens were collected from patients who were infected with Nocardia, and a part of the specimens were mixed with new sputum specimens from patients without Nocardia infection to enhance the data diversity. Secondly, the specimens were converted into smears with Gram staining. Images were captured under a microscope and subsequently annotated by experts, creating two datasets. Thirdly, each dataset was divided into three subsets: the training set, the validation set and the test set. The training and validation sets were used for training networks, while the test set was used for evaluating the effeteness of the trained networks. Finally, a neural network model was trained on this dataset, with an image of Gram-stained sputum smear as input, this model determines the presence and locations of Nocardia instances within the image.

Results

After training, the detection network was evaluated on two datasets, resulting in classification accuracies of 97.3% and 98.3%, respectively. This network can identify Nocardia instances in about 24 milliseconds per image on a personal computer. The detection metrics of mAP50 on both datasets were 0.780 and 0.841, respectively.

Conclusion

The Nocardia screening method can accurately and efficiently determine whether Nocardia exists in the images of Gram-stained sputum smears. Additionally, it can precisely locate the Nocardia instances, assisting doctors in confirming the presence of Nocardia.

Nano-Biotechnology for Bacteria Identification and Potent Anti-bacterial Properties: A Review of Current State of the Art

Sepsis is a critical disease caused by the abrupt increase of bacteria in human blood, which subsequently causes a cytokine storm. Early identification of bacteria is critical to treating a patient with proper antibiotics to avoid sepsis. However, conventional culture-based identification takes a long time. Polymerase chain reaction (PCR) is not so successful because of the complexity and similarity in the genome sequence of some bacterial species, making it difficult to design primers and thus less suitable for rapid bacterial identification. To address these issues, several new technologies have been developed. Recent advances in nanotechnology have shown great potential for fast and accurate bacterial identification. The most promising strategy in nanotechnology involves the use of nanoparticles, which has led to the advancement of highly specific and sensitive biosensors capable of detecting and identifying bacteria even at low concentrations in very little time. The primary drawback of conventional antibiotics is the potential for antimicrobial resistance, which can lead to the development of superbacteria, making them difficult to treat. The incorporation of diverse nanomaterials and designs of nanomaterials has been utilized to kill bacteria efficiently. Nanomaterials with distinct physicochemical properties, such as optical and magnetic properties, including plasmonic and magnetic nanoparticles, have been extensively studied for their potential to efficiently kill bacteria. In this review, we are emphasizing the recent advances in nano-biotechnologies for bacterial identification and anti-bacterial properties. The basic principles of new technologies, as well as their future challenges, have been discussed.

Label-free detection of trace level zearalenone in corn oil by surface-enhanced Raman spectroscopy (SERS) coupled with deep learning models

Surface-enhanced Raman spectroscopy (SERS) and deep learning models were adopted for detecting zearalenone (ZEN) in corn oil. First, gold nanorods were synthesized as a SERS substrate. Second, the collected SERS spectra were augmented to improve the generalization ability of regression models. Third, five regression models, including partial least squares regression (PLSR), random forest regression (RFR), Gaussian progress regression (GPR), one-dimensional convolutional neural networks (1D CNN), and two-dimensional convolutional neural networks (2D CNN), were developed. The results showed that 1D CNN and 2D CNN models possessed the best prediction performance, i.e., determination of prediction set (R_P²) = 0.9863 and 0.9872, root mean squared error of prediction set (RMSEP) = 0.2267 and 0.2341, ratio of performance to deviation (RPD) = 6.548 and 6.827, limit of detection (LOD) = 6.81 × 10^-4 and 7.24 × 10^-4 μg/mL. Therefore, the proposed method offers an ultrasensitive and effective strategy for detecting ZEN in corn oil.

Current strategies of plasmonic nanoparticles assisted surface-enhanced Raman scattering toward biosensor studies

With the progressive nanofabrication technology, plasmonic nanoparticles (PNPs) have been increasingly deployed in the field of biosensing. PNPs have favorable biocompatibility, conductivity, and tunable optical properties. In addition, the localized surface plasmon resonance (LSPR) of PNPs plays a vital role in surface-enhanced Raman scattering (SERS). PNPs-based SERS biosensing enables wide-ranging applications for sensitive detection and high spatial and temporal resolution imaging. Numerous reviews of PNPs in the field of SERS biosensing highlight the fabrication or applications in one or more fields. However, the specific strategies for the SERS biosensor construction had not been summarized systematically. Thus, this work offers a comprehensive overview of SERS enhancement strategies based on PNPs, with a focus on SERS label-free detection along with label detection sensing construction, as well as its challenges and future trends.

Simultaneous quantitative analysis of Escherichia coli, Staphylococcus aureus and Salmonella typhimurium using surface-enhanced Raman spectroscopy coupled with partial least squares regression and artificial neural networks

Simultaneous detection of mixed bacteria accurately and sensitively is a major challenge in microbial quality control field. In this study, we proposed a label-free SERS technique coupled with partial least squares regression (PLSR) and artificial neural networks (ANNs) for quantitative analysis of Escherichia coli, Staphylococcus aureus and Salmonella typhimurium simultaneously. SERS-active and reproducible Raman spectra can be acquired directly upon the bacteria and Au@Ag@SiO₂ nanoparticle composites on the surface of gold foil substrates. After applying different preprocessing models, SERS-PLSR and SERS-ANNs quantitative analysis models were developed to map SERS spectra of concentrations of the Escherichia coli, Staphylococcus aureus and Salmonella typhimurium, respectively. Both models achieved high prediction accuracy and low prediction error, while the performance of SERS-ANNs model in both quality of fit (R² > 0.95) and accuracy of predictions (RMSE < 0.06) was superior to SERS-PLSR model. Therefore, it is feasible to develop simultaneous quantitative analysis of mixed pathogenic bacteria by proposed SERS methodology.

Recent advances of Raman spectroscopy for the analysis of bacteria

Rapid and sensitive bacteria detection and identification are becoming increasingly important for a wide range of areas including the control of food safety, the prevention of infectious diseases, and environmental monitoring. Raman spectroscopy is an emerging technology which provides comprehensive information for the analysis of bacteria in a short time and with high sensitivity. Raman spectroscopy offers many advantages including relatively simple operation, non-destructive analysis, and information on molecular differences between bacteria species and strains. A variety of biochemical properties can be measured in a single spectrum. This short review covers the recent advancements and applications of Raman spectroscopy for bacteria analysis with specific focuses on bacteria detection, bacteria identification and discrimination, as well as bacteria antibiotic susceptibility testing in 2022. The development of novel substrates, the combination with other techniques, and the utilization of advanced data processing tools for the improvement of Raman spectroscopy and future directions are discussed.

Discovering the digital biomarker of hepatocellular carcinoma in serum with SERS-based biosensors and intelligence vision

By its many virtues, non-biomarker-reliant molecular detection has recently shown bright prospects for cancer screening but its clinical application is hindered by the shortage of measurable criteria that are analogous to biomarkers. Here, we report a digital biomarker, as a new-concept serum biomarker, of hepatocellular carcinoma (HCC) found with SERS-based biosensors and a deep neural network "digital retina" for visualizing and explicitly defining spectral fingerprints. We validate the discovered digital biomarker (a collection of 10 characteristic peaks in the serum SERS spectra) with unsupervised clustering of spectra from an independent sample batch comprised normal individuals and HCC cases; the validation results show clustering accuracies of 95.71% and 100.00%, respectively. Furthermore, we find that the digital biomarker of HCC shares a few common peaks with three clinically applied serum biomarkers, which means it could convey essential biomolecular information similar to these biomarkers. Accordingly, we present an intelligent method for early HCC detection that leverages the digital biomarker with similar traits as biomarkers. Employing the digital biomarker, we could accurately stratify HCC, hepatitis B, and normal populations with linear classifiers, exhibiting accuracies over 92% and area under the receiver operating curve values above 0.93. It is anticipated that this non-biomarker-reliant molecular detection method will facilitate mass cancer screening.

Identification of antibiotic residues in aquatic products with surface-enhanced Raman scattering powered by 1-D convolutional neural networks

Universal and fast antibiotic residues detection technology is imperative for the control of food safety in aquatic products. However, accurate surface-enhanced Raman scattering (SERS) quantitative detection of complicated samples is still a challenge. A recognition method powered by deep learning and took advantage of the unique fingerprint information merits of SERS was proposed. Herein, the spectra were collected by Ag nanofilm SERS substrate prepared by self-assembly of Ag nanoparticles on water/oil interface. A SERS-based database of commonly used antibiotics in aquatic products was set up, which is suitable for employed as input data for learning and training. The results show that the five types of antibiotics are successfully distinguished through principal component analysis (PCA) and each antibiotic in every type was successfully distinguished. Furthermore, one-dimensional convolutional neural networks (1-D CNN) was used to distinguish the antibiotics, and the results show that all the test samples were correctly predicted by 1-D CNN model. The results of this research suggest the great potential of the combination of SERS spectra with deep learning as a method for rapid and highly accurate identification of antibiotic residues in aquatic products.

Single-cell pathogen diagnostics for combating antibiotic resistance

Bacterial infections and antimicrobial resistance are a major cause for morbidity and mortality worldwide. Antimicrobial resistance often arises from antimicrobial misuse, where physicians empirically treat suspected bacterial infections with broad-spectrum antibiotics until standard culture-based diagnostic tests can be completed. There has been a tremendous effort to develop rapid diagnostics in support of the transition from empirical treatment of bacterial infections towards a more precise and personalized approach. Single-cell pathogen diagnostics hold particular promise, enabling unprecedented quantitative precision and rapid turnaround times. This Primer provides a guide for assessing, designing, implementing and applying single-cell pathogen diagnostics. First, single-cell pathogen diagnostic platforms are introduced based on three essential capabilities: cell isolation, detection assay and output measurement. Representative results, common analysis methods and key applications are highlighted, with an emphasis on initial screening of bacterial infection, bacterial species identification and antimicrobial susceptibility testing. Finally, the limitations of existing platforms are discussed, with perspectives offered and an outlook towards clinical deployment. This Primer hopes to inspire and propel new platforms that can realize the vision of precise and personalized bacterial infection treatments in the near future.

Nanomaterials meet surface-enhanced Raman scattering towards enhanced clinical diagnosis: a review

Surface-enhanced Raman scattering (SERS) is a very promising tool for the direct detection of biomarkers for the diagnosis of i.e., cancer and pathogens. Yet, current SERS strategies are hampered by non-specific interactions with co-existing substances in the biological matrices and the difficulties of obtaining molecular fingerprint information from the complex vibrational spectrum. Raman signal enhancement is necessary, along with convenient surface modification and machine-based learning to address the former issues. This review aims to describe recent advances and prospects in SERS-based approaches for cancer and pathogens diagnosis. First, direct SERS strategies for key biomarker sensing, including the use of substrates such as plasmonic, semiconductor structures, and 3D order nanostructures for signal enhancement will be discussed. Secondly, we will illustrate recent advances for indirect diagnosis using active nanomaterials, Raman reporters, and specific capture elements as SERS tags. Thirdly, critical challenges for translating the potential of the SERS sensing techniques into clinical applications via machine learning and portable instrumentation will be described. The unique nature and integrated sensing capabilities of SERS provide great promise for early cancer diagnosis or fast pathogens detection, reducing sanitary costs but most importantly allowing disease prevention and decreasing mortality rates.

Label-Free Surface-Enhanced Raman Spectroscopic Analysis of Proteins: Advances and Applications

Surface-enhanced Raman spectroscopy (SERS) is powerful for structural characterization of biomolecules under physiological condition. Owing to its high sensitivity and selectivity, SERS is useful for probing intrinsic structural information of proteins and is attracting increasing attention in biophysics, bioanalytical chemistry, and biomedicine. This review starts with a brief introduction of SERS theories and SERS methodology of protein structural characterization. SERS-active materials, related synthetic approaches, and strategies for protein-material assemblies are outlined and discussed, followed by detailed discussion of SERS spectroscopy of proteins with and without cofactors. Recent applications and advances of protein SERS in biomarker detection, cell analysis, and pathogen discrimination are then highlighted, and the spectral reproducibility and limitations are critically discussed. The review ends with a conclusion and a discussion of current challenges and perspectives of promising directions.

Escherichia coli Enumeration in a Capillary-Driven Microfluidic Chip with SERS

Pathogen detection is still a challenging issue for public health, especially in food products. A selective preconcentration step is also necessary if the target pathogen concentration is very low or if the sample volume is limited in the analysis. Plate counting (24-48 h) methods should be replaced by novel biosensor systems as an alternative reliable pathogen detection technique. The usage of a capillary-driven microfluidic chip is an alternative method for pathogen detection, with the combination of surface-enhanced Raman scattering (SERS) measurements. Here, we constructed microchambers with capillary microchannels to provide nanoparticle-pathogen transportation from one chamber to the other. Escherichia coli (E. coli) was selected as a model pathogen and specific antibody-modified magnetic nanoparticles (MNPs) as a capture probe in a complex milk matrix. MNPs that captured E. coli were transferred in a capillary-driven microfluidic chip consisting of four chambers, and 4-aminothiophenol (4-ATP)-labelled gold nanorods (Au NRs) were used as the Raman probe in the capillary-driven microfluidic chip. The MNPs provided immunomagnetic (IMS) separation and preconcentration of analytes from the sample matrix and then, 4-ATP-labelled Au NRs provided an SERS response by forming sandwich immunoassay structures in the last chamber of the capillary-driven microfluidic chip. The developed SERS-based method could detect 101-107 cfu/mL of E. coli with the total analysis time of less than 60 min. Selectivity of the developed method was also tested by using Salmonella enteritidis (S. enteritidis) and Staphylococcus aureus (S. aureus) as analytes, and very weak signals were observed.

Specific and quantitative detection of bacteria based on surface cell imprinted SERS mapping platform

Non-specificity and poor quantitative ability are the main challenges in surface-enhanced Raman scattering (SERS) technique, especially for the detection of bacteria in real samples. In this study, we presented a surface cell imprinted SERS mapping platform which is competent for the specific and quantitative detection of bacteria. The platform based on the fabrication of a surface cell imprinted substrate (SCIS) by which Escherichia coli (E. coli) can be captured and labelled by SERS tags which produces strong characteristic signal to indicate the capture of targets. We highlighted the specificity of this platform in the detection of E. coli, by comparing the performances toward Salmonella paratyphoid A, Bacillus subtilis, Enterococcus faecalis and Staphylococcus aureus. Upon integrating with SERS mapping technique, the platform displayed good quantitative ability toward E. coli with a wide linear range from 10² to 10⁸ CFU/mL and a low detection limit of ∼1.35 CFU/mL. Moreover, this novel SERS analysis platform was proved to be effective for E. coli detection in real probiotic beverage and chicken breast meat samples. By fabricating different SCISs, this platform can be replicated for the detection of other bacteria, which provides a promising application for real sample testing.

Reusable and universal impedimetric sensing platform for the rapid and sensitive detection of pathogenic bacteria based on bacteria-imprinted polythiophene film

A bacteria-imprinted polythiophene film (BIF)-based impedimetric sensor was proposed for the rapid and sensitive detection of S. aureus. A significant improvement is the reduced time for both BIF fabrication (15 min) and bacterial capturing (10 min).