Early detection and classification of live bacteria using time-lapse coherent imaging and deep learning

Deep learning-based detection of bacterial swarm motion using a single image

Motility is a fundamental characteristic of bacteria. Distinguishing between swarming and swimming, the two principal forms of bacterial movement, holds significant conceptual and clinical relevance. Conventionally, the detection of bacterial swarming involves inoculating samples on an agar surface and observing colony expansion, which is qualitative, time-intensive, and requires additional testing to rule out other motility forms. A recent methodology that differentiates swarming and swimming motility in bacteria using circular confinement offers a rapid approach to detecting swarming. However, it still heavily depends on the observer's expertise, making the process labor-intensive, costly, slow, and susceptible to inevitable human bias. To address these limitations, we developed a deep learning-based swarming classifier that rapidly and autonomously predicts swarming probability using a single blurry image. Compared with traditional video-based, manually processed approaches, our method is particularly suited for high-throughput environments and provides objective, quantitative assessments of swarming probability. The swarming classifier demonstrated in our work was trained on Enterobacter sp. SM3 and showed good performance when blindly tested on new swarming (positive) and swimming (negative) test images of SM3, achieving a sensitivity of 97.44% and a specificity of 100%. Furthermore, this classifier demonstrated robust external generalization capabilities when applied to unseen bacterial species, such as Serratia marcescens DB10 and Citrobacter koseri H6. This competitive performance indicates the potential to adapt our approach for diagnostic applications through portable devices, which would facilitate rapid, objective, on-site screening for bacterial swarming motility, potentially enhancing the early detection and treatment assessment of various diseases, including inflammatory bowel diseases (IBD) and urinary tract infections (UTI).

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.

Engineered Artificial Nanochannels with Cell Membrane Nanointerface for Ultrasensitive Detection and Discrimination of Multiple Bacterial Infections

Bacterial infection is a major threat to global public health, which urgently require rapid and reliable analytical techniques for complex biological samples but remains a challenge. Herein, we developed an artificial affinity nanochannel with a cell membrane nanointerface, which enables broad-spectrum capture and specifically discriminates multiple pathogens. The macrophage membrane is pre-engineered with azide groups by a biometabolic process and then modified on a porous anodized aluminum oxide substrate via click reactions, preserving dynamic lateral fluidity and broad-spectrum recognition capacity. The macrophage membrane/anodized aluminum oxide membrane is evaluated with remarkable ion current rectification performance with a distinct current response upon bacterial binding, which realizes ultrasensitive detection of bacteria. Moreover, discrimination of bacterial species is achieved by further introducing specific antibodies. The nanochannel-based biosensor allows accurately capturing and quantifying multiple bacteria over a broad linear range, with a detection limit as low as 2.7 CFU/mL. Finally, this nanoplatform is successfully applied for broad-spectrum capture of bacterial species in several practical application scenarios including water, serum, and blood samples, achieving ultrasensitive detection and identification of bacteria below 10 CFU/mL. Overall, the proposed nanochannel with cell membrane nanointerface shows broad applicability in bacterial analysis, highlighting its potential in diagnosing infectious diseases.

Machine learning in point-of-care testing: innovations, challenges, and opportunities

The landscape of diagnostic testing is undergoing a significant transformation, driven by the integration of artificial intelligence (AI) and machine learning (ML) into decentralized, rapid, and accessible sensor platforms for point-of-care testing (POCT). The COVID-19 pandemic has accelerated the shift from centralized laboratory testing but also catalyzed the development of next-generation POCT platforms that leverage ML to enhance the accuracy, sensitivity, and overall efficiency of point-of-care sensors. This Perspective explores how ML is being embedded into various POCT modalities, including lateral flow assays, vertical flow assays, nucleic acid amplification tests, and imaging-based sensors, illustrating their impact through different applications. We also discuss several challenges, such as regulatory hurdles, reliability, and privacy concerns, that must be overcome for the widespread adoption of ML-enhanced POCT in clinical settings and provide a comprehensive overview of the current state of ML-driven POCT technologies, highlighting their potential impact in the future of healthcare.

Initializing a Public Repository for Hosting Benchmark Datasets to Facilitate Machine Learning Model Development in Food Safety

While there is clear potential for artificial intelligence (AI) and machine learning (ML) models to help improve food safety, the development and deployment of these models in the food safety domain are by and large lacking. The absence of publicly available databases that host well-curated datasets that can be used to develop and validate AI /ML models represents one likely barrier. Thus, we took three previously published datasets, which we further cleaned and annotated, and made them publicly available in a repository called Cornell Food Safety ML Repository. The selected datasets include (i) presence or absence of Listeria spp. in soil samples collected across the U.S. with paired metadata for soil properties, geolocation, climate, and surrounding land use, (ii) presence or absence of Salmonella and Campylobacter in young chicken carcasses tested in processing facilities with associated meteorological and temporal metadata, and (iii) presence or absence of fecal contamination as well as E. coli concentration in New York watersheds with associated metadata for land use, water attributes, and meteorological factors. These datasets can serve as benchmark datasets for developing ML models. To demonstrate the utility of the repository, we developed customizable scripts as well as LazyPredict (a quick screening method) scripts for training different types of ML models using the shared datasets. While this repository provides an important starting point that will allow for the development and testing of ML models to predict foodborne pathogens contamination in different sources, the inclusion of further datasets is clearly needed to advance this field. This paper thus includes a call to action for the deposit of well-curated datasets that can be used for further development of predictive models in food safety. This paper will also discuss the benefits of such public databases, including the assessment of data-sharing scenarios using existing privacy-preserving techniques.

Spectroscopic characterization of bacterial colonies through UV hyperspectral imaging techniques

Introduction

Plate culturing and visual inspection are the gold standard methods for bacterial identification. Despite the growing attention on molecular biology techniques, colony identification using agar plates remains manual, interpretative, and heavily reliant on human experience, making it prone to errors. Advanced imaging techniques, like hyperspectral imaging, offer potential alternatives. However, the use of hyperspectral imaging in the VIS-NIR region has been hindered by sensitivity to various components and culture medium changes, leading to inaccurate results. The application of hyperspectral imaging in the ultraviolet (UV) region has not been explored, despite the presence of specific absorption and emission peaks in bacterial components.

Methods

To address this gap, we developed a predictive model for bacterial colony detection and identification using UV hyperspectral imaging. The model utilizes hyperspectral images acquired in the UV wavelength range of 225-400 nm, processed with principal component analysis (PCA) and discriminant analysis (DA). The measurement setup includes a hyperspectral imager, a PC for automated data analysis, and a conveyor belt system to transport agar plates for automated analysis. Four bacterial species (Escherichia coli, Staphylococcus, Pseudomonas, and Shewanella) were cultured on two different media, Luria Bertani and Tryptic Soy, to train and validate the model.

Results

The PCA-DA-based model demonstrated high accuracy (90%) in differentiating bacterial species based on the first three principal components, highlighting the potential of UV hyperspectral imaging for bacterial identification.

Discussion

This study shows that UV hyperspectral imaging, coupled with advanced data analysis techniques, offers a robust and automated alternative to traditional methods for bacterial identification. The model's high accuracy emphasizes the untapped potential of UV hyperspectral imaging in microbiological analysis, reducing human error and improving reliability in bacterial species differentiation.

High-content imaging and deep learning-driven detection of infectious bacteria in wounds

Fast and accurate detection of infectious bacteria in wounds is crucial for effective clinical treatment. However, traditional methods take over 24 h to yield results, which is inadequate for urgent clinical needs. Here, we introduce a deep learning-driven framework that detects and classifies four bacteria commonly found in wounds: Acinetobacter baumannii (AB), Escherichia coli (EC), Pseudomonas aeruginosa (PA), and Staphylococcus aureus (SA). This framework leverages the pretrained ResNet50 deep learning architecture, trained on manually collected periodic bacterial colony-growth images from high-content imaging. In in vitro samples, our method achieves a detection rate of over 95% for early colonies cultured for 8 h, reducing detection time by more than 12 h compared to traditional Environmental Protection Agency (EPA)-approved methods. For colony classification, it identifies AB, EC, PA, and SA colonies with accuracies of 96%, 97%, 96%, and 98%, respectively. For mixed bacterial samples, it identifies colonies with 95% accuracy and classifies them with 93% precision. In mouse wound samples, the method identifies over 90% of developing bacterial colonies and classifies colony types with an average accuracy of over 94%. These results highlight the framework's potential for improving the clinical treatment of wound infections. Besides, the framework provides the detection results with key feature visualization, which enhance the prediction credibility for users. To summarize, the proposed framework enables high-throughput identification, significantly reducing detection time and providing a cost-effective tool for early bacterial detection.

Optical time-lapse microscopy for rapid assessment of microbial quality in hygroscopic food samples

In the food industry, time-to-result is crucial for faster release of products, minimising recalls, mitigation of microbial contamination problems and, ultimately, food safety. Carrageenan is isolated from red seaweed (Rhodophyta) and applied in various foods and beverages as a gelling, thickening, texturing, or stabilising agent due to its hygroscopic properties. Currently, the standard industry plate count method entails a one-hundred-fold dilution of the sample before mixing with molten agar for assessing the level of microbial contamination in carrageenan samples prior to business-to-business shipment. However, even at this dilution, carrageenan swells, forms clumps, clogs pipettes, and leaves thick gel structures, bubbles, and debris in agar plates causing microbial enumeration to be challenging and subject to human error. Here, we report, for the first time, the application of mini agar plates monitored by the automated time-lapse microscopy IntuGrow solution to assess the microbiological quality in the challenging food ingredient. Without dilution of the food sample, the carrageenan powder is scattered between two layers of Plate Count Agar to enumerate bacteria within 12-20 h, while enumeration by traditional plate counts requires 72 h. A DELAY algorithm for optical time-lapse microscopy was developed and added to IntuGrow analysis software to suppress the effects of swelling and enhance detection of the presence of growing microbial colonies by normalising the background using previous images. Time-lapse microscopy image-based monitoring made it possible to obtain results from carrageenan samples that could not be obtained by traditional plate counts due to swarming bacteria. Comparison between the two methods showed a nearly perfect Demings slope of 0.96, while an observed bias of -0.33 log CFU/g indicated that IntuGrow counts were lower than traditional plate counts. This is likely due to carrageenan artefacts being counted as colonies in the latter plates. The ability of IntuGrow to enumerate bacteria in challenging food ingredients such as carrageenan implies that the technology should be easy to apply for easy-to-dilute samples or non-hydrocolloid powders. Further testing in an industrial setting by different operators should be used to validate the reproducibility of the method.

Self-Interference Digital Optofluidic Genotyping for Integrated and Automated Label-Free Pathogen Detection

Pathogen, prevalent in both natural and human environments, cause approximately 4.95 million deaths annually, ranking them among the top contributors to global mortality. Traditional pathogen detection methods, reliant on microscopy and cultivation, are slow and labor-intensive and often produce subjective results. While nucleic acid amplification techniques such as polymerase chain reaction offer genetic accuracy, they necessitate costly laboratory equipment and skilled personnel. Consequently, isothermal amplification methods like recombinase polymerase amplification (RPA) have attracted interest for their rapid and straightforward operations. However, these methods face challenges in specificity and automated sample processing. In this study, we introduce a self-interferometric digital optofluidic platform incorporating asymmetric direct solid-phase RPA for real-time, label-free, and automated pathogen genotyping. By integration of digital microfluidics with a DNA monolayer detection method using hyperspectral interferometry, this platform enables rapid, specific, and sensitive pathogen detection without the need for exogenous labeling or complex procedures. The system demonstrated high sensitivity (10 CFU·mL-1), specificity (differentiating four Candida species), detection efficiency (fully automated within 50 min for Gram-negative bacteria), and throughput (simultaneous detection of four indices). This integrated approach to pathogen quantitation on a single microfluidic chip represents a significant advancement in rapid pathogen diagnostics, providing a practical solution for timely pathogen detection and analysis.

Utilization of AI - reshaping the future of food safety, agriculture and food security - a critical review

Artificial intelligence is an emerging technology which harbors a suite of mechanisms that have the potential to be leveraged for reaping value across multiple domains. Lately, there is an increased interest in embracing applications associated with Artificial Intelligence to positively contribute to food safety. These applications such as machine learning, computer vision, predictive analytics algorithms, sensor networks, robotic inspection systems, and supply chain optimization tools have been established to contribute to several domains of food safety such as early warning of outbreaks, risk prediction, detection and identification of food associated pathogens. Simultaneously, the ambition toward establishing a sustainable food system has motivated the adoption of cutting-edge technologies such as Artificial Intelligence to strengthen food security. Given the myriad challenges confronting stakeholders in their endeavors to safeguard food security, Artificial Intelligence emerges as a promising tool capable of crafting holistic management strategies for food security. This entails maximizing crop yields, mitigating losses, and trimming operational expenses. AI models present notable benefits in efficiency, precision, uniformity, automation, pattern identification, accessibility, and scalability for food security endeavors. The escalation in the global trend for adopting alternative protein sources such as edible insects and microalgae as a sustainable food source reflects a growing recognition of the need for sustainable and resilient food systems to address the challenges of population growth, environmental degradation, and food insecurity. Artificial Intelligence offers a range of capabilities to enhance food safety in the production and consumption of alternative proteins like microalgae and edible insects, contributing to a sustainable and secure food system.

Simple Imaging System for Label-Free Identification of Bacterial Pathogens in Resource-Limited Settings

Fast, accurate, and affordable bacterial identification methods are paramount for the timely treatment of infections, especially in resource-limited settings (RLS). However, today, only 1.3% of the sub-Saharan African diagnostic laboratories are performing clinical bacteriology. To improve this, diagnostic tools for RLS should prioritize simplicity, affordability, and ease of maintenance, as opposed to the costly equipment utilized for bacterial identification in high-income countries, such as matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS). In this work, we present a new high-throughput approach based on a simple wide-field (864 mm2) lensless imaging system allowing for the acquisition of a large portion of a Petri dish coupled with a supervised deep learning algorithm for identification at the bacterial colony scale. This wide-field imaging system is particularly well suited to RLS since it includes neither moving mechanical parts nor optics. We validated this approach through the acquisition and the subsequent analysis of a dataset comprising 252 clinical isolates from five species, encompassing some of the most prevalent pathogens. The resulting optical morphotypes exhibited intra- and interspecies variability, a scenario considerably more akin to real-world clinical practice than the one achievable by solely concentrating on reference strains. Despite this variability, high identification performance was achieved with a correct species identification rate of 91.7%. These results open up some new prospects for identification in RLS. We released both the acquired dataset and the trained identification algorithm in publicly available repositories.

High-throughput fluorescence sensing array based on tetraphenylethylene derivatives for detecting and distinguishing pathogenic microbes

Infections induced by pathogenic microorganisms will bring negative effects such as diseases that damage health and result in heavy economic burden. Therefore, it is very important to detect and identify the pathogens in time. Moreover, traditional clinical diagnosis or food testing often faces the problem of dealing with a large number of samples. Here, we designed a high-throughput fluorescent sensor array based on the different binding ability of five tetraphenylethylene derivatives (TPEs) with various side chains to different kinds of pathogenic microbes, which is used to detect and distinguish various species, so as to realize rapid mass diagnosis, and hopefully provide guidance for further determination of microbial infections and clinical treatment.

Application of machine learning based genome sequence analysis in pathogen identification

Infectious diseases caused by pathogenic microorganisms pose a serious threat to human health. Despite advances in molecular biology, genetics, computation, and medicinal chemistry, infectious diseases remain a significant public health concern. Addressing the challenges posed by pathogen outbreaks, pandemics, and antimicrobial resistance requires concerted interdisciplinary efforts. With the development of computer technology and the continuous exploration of artificial intelligence(AI)applications in the biomedical field, the automatic morphological recognition and image processing of microbial images under microscopes have advanced rapidly. The research team of Institute of Microbiology, Chinese Academy of Sciences has developed a single cell microbial identification technology combining Raman spectroscopy and artificial intelligence. Through laser Raman acquisition system and convolutional neural network analysis, the average accuracy rate of 95.64% has been achieved, and the identification can be completed in only 5 min. These technologies have shown substantial advantages in the visible morphological detection of pathogenic microorganisms, expanding anti-infective drug discovery, enhancing our understanding of infection biology, and accelerating the development of diagnostics. In this review, we discuss the application of AI-based machine learning in image analysis, genome sequencing data analysis, and natural language processing (NLP) for pathogen identification, highlighting the significant role of artificial intelligence in pathogen diagnosis. AI can improve the accuracy and efficiency of diagnosis, promote early detection and personalized treatment, and enhance public health safety.

Polydopamine modified colloidal gold nanotag-based lateral flow immunoassay platform for highly sensitive detection of pathogenic bacteria and fast evaluation of antibacterial agents

Bacterial infection is a great threat to human health. Lateral flow immunoassays (LFIAs) with the merits of low cost, quick screening, and on-site detection are competitive technologies for bacteria detection, but their detection limits depend on the optical performance of the adopted nanotags. Herein, we presented a LFIA platform for bacteria detection using polydopamine (PDA) functionalized Au nanoparticles (denoted as Au@PDA) as the nanotag. The introduction of PDA could provide enhanced light absorption of Au, as well as numerous functional groups for conjugation. Small recognition molecules i.e. vancomycin (Van) and p-mercaptophenylboronic acid (PMBA) were covalently anchored to Au@PDA, and selected as the specific probes towards Gram-positive (G+) and Gram-negative (G-) bacteria, respectively. Taken Staphylococcus aureus (S. aureus) and Escherichia coli (E. coli) as the representative targets of G+ and G- bacteria, two LFA strips were successfully constructed based on the immuno-sandwich principle. They could quantitatively detect S. aureus and E. coli both down to 102 cfu/mL, a very competitive detection limit in comparison with other colorimetric or luminescent probes-based LFIAs. Furthermore, the proposed two strips were applied for the quantitative, accurate, and rapid detection of S. aureus and E. coli in food and human urine samples with good analytical results obtained. In addition, they were integrated as a screening platform for quick evaluation of diverse antibacterial agents within 3 h, which is remarkably shortened compared with that of the two traditional methods i.e. bacterial culture and plate-counting.

Roadmap on computational methods in optical imaging and holography [invited]

Computational methods have been established as cornerstones in optical imaging and holography in recent years. Every year, the dependence of optical imaging and holography on computational methods is increasing significantly to the extent that optical methods and components are being completely and efficiently replaced with computational methods at low cost. This roadmap reviews the current scenario in four major areas namely incoherent digital holography, quantitative phase imaging, imaging through scattering layers, and super-resolution imaging. In addition to registering the perspectives of the modern-day architects of the above research areas, the roadmap also reports some of the latest studies on the topic. Computational codes and pseudocodes are presented for computational methods in a plug-and-play fashion for readers to not only read and understand but also practice the latest algorithms with their data. We believe that this roadmap will be a valuable tool for analyzing the current trends in computational methods to predict and prepare the future of computational methods in optical imaging and holography.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00340-024-08280-3.

Artificial intelligence applications in the diagnosis and treatment of bacterial infections

The diagnosis and treatment of bacterial infections in the medical and public health field in the 21st century remain significantly challenging. Artificial Intelligence (AI) has emerged as a powerful new tool in diagnosing and treating bacterial infections. AI is rapidly revolutionizing epidemiological studies of infectious diseases, providing effective early warning, prevention, and control of outbreaks. Machine learning models provide a highly flexible way to simulate and predict the complex mechanisms of pathogen-host interactions, which is crucial for a comprehensive understanding of the nature of diseases. Machine learning-based pathogen identification technology and antimicrobial drug susceptibility testing break through the limitations of traditional methods, significantly shorten the time from sample collection to the determination of result, and greatly improve the speed and accuracy of laboratory testing. In addition, AI technology application in treating bacterial infections, particularly in the research and development of drugs and vaccines, and the application of innovative therapies such as bacteriophage, provides new strategies for improving therapy and curbing bacterial resistance. Although AI has a broad application prospect in diagnosing and treating bacterial infections, significant challenges remain in data quality and quantity, model interpretability, clinical integration, and patient privacy protection. To overcome these challenges and, realize widespread application in clinical practice, interdisciplinary cooperation, technology innovation, and policy support are essential components of the joint efforts required. In summary, with continuous advancements and in-depth application of AI technology, AI will enable doctors to more effectivelyaddress the challenge of bacterial infection, promoting the development of medical practice toward precision, efficiency, and personalization; optimizing the best nursing and treatment plans for patients; and providing strong support for public health safety.

Starting with screening strains to construct synthetic microbial communities (SynComs) for traditional food fermentation

With the elucidation of community structures and assembly mechanisms in various fermented foods, core communities that significantly influence or guide fermentation have been pinpointed and used for exogenous restructuring into synthetic microbial communities (SynComs). These SynComs simulate ecological systems or function as adjuncts or substitutes in starters, and their efficacy has been widely verified. However, screening and assembly are still the main limiting factors for implementing theoretic SynComs, as desired strains cannot be effectively obtained and integrated. To expand strain screening methods suitable for SynComs in food fermentation, this review summarizes the recent research trends in using SynComs to study community evolution or interaction and improve the quality of food fermentation, as well as the specific process of constructing synthetic communities. The potential for novel screening modalities based on genes, enzymes and metabolites in food microbial screening is discussed, along with the emphasis on strategies to optimize assembly for facilitating the development of synthetic communities.

Apt-Conjugated PDMS-ZnO/Ag-Based Multifunctional Integrated Superhydrophobic Biosensor with High SERS Activity and Photocatalytic Sterilization Performance

Sensitive detection and efficient inactivation of pathogenic bacteria are crucial for halting the spread and reproduction of foodborne pathogenic bacteria. Herein, a novel Apt-modified PDMS-ZnO/Ag multifunctional biosensor has been developed for high-sensitivity surface-enhanced Raman scattering (SERS) detection along with photocatalytic sterilization towards Salmonella typhimurium (S. typhimurium). The distribution of the electric field in PDMS-ZnO/Ag with different Ag sputtering times was analyzed using a finite-difference time-domain (FDTD) algorithm. Due to the combined effect of electromagnetic enhancement and chemical enhancement, PDMS-ZnO/Ag exhibited outstanding SERS sensitivity. The limit of detection (LOD) for 4-MBA on the optimal SERS substrate (PZA-40) could be as little as 10-9 M. After PZA-40 was modified with the aptamer, the LOD of the PZA-40-Apt biosensor for detecting S. typhimurium was only 10 cfu/mL. Additionally, the PZA-40-Apt biosensor could effectively inactivate S. typhimurium under visible light irradiation within 10 min, with a bacterial lethality rate (Lb) of up to 97%. In particular, the PZA-40-Apt biosensor could identify S. typhimurium in food samples in addition to having minimal cytotoxicity and powerful biocompatibility. This work provides a multifunctional nanoplatform with broad prospects for selective SERS detection and photocatalytic sterilization of pathogenic bacteria.

Periodontal disease and emerging point-of-care technologies for its diagnosis

Periodontal disease (PD), a chronic inflammatory disorder that damages the tooth and its supporting components, is a common global oral health problem. Understanding the intricacies of these disorders, from gingivitis to severe PD, is critical for efficient treatment, diagnosis, and prevention in dental care. Periodontal biosensors and biomarkers are critical in improving oral health diagnostic skills. Clinicians may accomplish early identification, tailored therapy, and efficient tracking of periodontal diseases by using these technologies, ushering in a new age of accurate oral healthcare. Traditional periodontitis diagnostic methods frequently rely on physical probing and visual examinations, necessitating the development of point-of-care (POC) devices. As periodontal disorders necessitate more precise and rapid diagnosis, incorporating novel innovations in biosensors and biomarkers becomes increasingly crucial. These innovations improve our capacity to diagnose, monitor, and adapt periodontal therapies, bringing in the next phase of customized and effective dental healthcare. The review discusses the characteristics and stages of PD, clinical treatment techniques, prominent biomarkers and infection-associated factors that may be employed to determine PD, biomedical sensing, and POC appliances that have been created so far to diagnose stages of PD and its progression profile, as well as predicting future developments in this field.

Modernization of digital food safety control

Foodborne illness remains a pressing global issue due to the complexities of modern food supply chains and the vast array of potential contaminants that can arise at every stage of food processing from farm to fork. Traditional food safety control systems are increasingly challenged to identify these intricate hazards. The U.S. Food and Drug Administration's (FDA) New Era of Smarter Food Safety represents a revolutionary shift in food safety methodology by leveraging cutting-edge digital technologies. Digital food safety control systems employ modern solutions to monitor food quality by efficiently detecting in real time a wide range of contaminants across diverse food matrices within a short timeframe. These systems also utilize digital tools for data analysis, providing highly predictive assessments of food safety risks. In addition, digital food safety systems can deliver a secure and reliable food supply chain with comprehensive traceability, safeguarding public health through innovative technological approaches. By utilizing new digital food safety methods, food safety authorities and businesses can establish an efficient regulatory framework that genuinely ensures food safety. These cutting-edge approaches, when applied throughout the food chain, enable the delivery of safe, contaminant-free food products to consumers.

All-optical complex field imaging using diffractive processors

Complex field imaging, which captures both the amplitude and phase information of input optical fields or objects, can offer rich structural insights into samples, such as their absorption and refractive index distributions. However, conventional image sensors are intensity-based and inherently lack the capability to directly measure the phase distribution of a field. This limitation can be overcome using interferometric or holographic methods, often supplemented by iterative phase retrieval algorithms, leading to a considerable increase in hardware complexity and computational demand. Here, we present a complex field imager design that enables snapshot imaging of both the amplitude and quantitative phase information of input fields using an intensity-based sensor array without any digital processing. Our design utilizes successive deep learning-optimized diffractive surfaces that are structured to collectively modulate the input complex field, forming two independent imaging channels that perform amplitude-to-amplitude and phase-to-intensity transformations between the input and output planes within a compact optical design, axially spanning ~100 wavelengths. The intensity distributions of the output fields at these two channels on the sensor plane directly correspond to the amplitude and quantitative phase profiles of the input complex field, eliminating the need for any digital image reconstruction algorithms. We experimentally validated the efficacy of our complex field diffractive imager designs through 3D-printed prototypes operating at the terahertz spectrum, with the output amplitude and phase channel images closely aligning with our numerical simulations. We envision that this complex field imager will have various applications in security, biomedical imaging, sensing and material science, among others.

Line image sensor-based colony fingerprinting system for rapid pathogenic bacteria identification

The rapid identification of pathogenic bacteria is crucial across various industries, including food or beverage manufacturing. Bacterial microcolony image-based classification has emerged as a promising approach to expedite identification, automate inspections, and reduce costs. However, conventional imaging methods have significant practical limitations, namely low throughput caused by the limited imaging range and slow imaging speed. To address these challenges, we developed an imaging system based on a line image sensor for rapid and wide-field imaging compared to existing colony imaging methods. This system can image a standard Petri dish (92 mm in diameter) completely within 22 s, successfully acquiring bacterial microcolony images. This process yielded a set of discrimination parameters termed as colony fingerprints, which were employed for machine learning. We demonstrated the performance of our system by identifying Staphylococcus aureus in food products using a machine learning model trained on a colony fingerprint dataset of 15 species from 9 genera, including foodborne pathogens. While conventional mass spectrometry-based methods require 24 h of incubation, our colony fingerprinting approach achieved 96% accuracy in just 10 h of incubation. Line image sensor offer high imaging speeds and scalability, allowing for swift and straightforward microbiological testing, eliminating the need for specialized expertise and overcoming the limitations of conventional methods. This innovation marks a transformative shift in industrial applications.

A high-speed microscopy system based on deep learning to detect yeast-like fungi cells in blood

Background: Blood-invasive fungal infections can cause the death of patients, while diagnosis of fungal infections is challenging. Methods: A high-speed microscopy detection system was constructed that included a microfluidic system, a microscope connected to a high-speed camera and a deep learning analysis section. Results: For training data, the sensitivity and specificity of the convolutional neural network model were 93.5% (92.7-94.2%) and 99.5% (99.1-99.5%), respectively. For validating data, the sensitivity and specificity were 81.3% (80.0-82.5%) and 99.4% (99.2-99.6%), respectively. Cryptococcal cells were found in 22.07% of blood samples. Conclusion: This high-speed microscopy system can analyze fungal pathogens in blood samples rapidly with high sensitivity and specificity and can help dramatically accelerate the diagnosis of fungal infectious diseases.

Effects of extreme precipitation on bacterial communities and bioaerosol composition: Dispersion in urban outdoor environments and health risks

Concerns about contaminants dispersed by seasonal precipitation have grown due to their potential hazards to outdoor environments and human health. However, studies on the crucial environmental factors influencing dispersion changes in bacterial communities are limited. This research adopted four-season in situ monitoring and sequencing techniques to examine the regional distribution profiles of bioaerosols, bacterial communities, and risks associated with extreme snowfall versus rainfall events in two monsoon cities. In the early-hours of winter snowfall, airborne cultivable bioaerosol concentrations were 4.1 times higher than the reference exposure limit (500 CFU/m3). The concentration of ambient particles (2.5 μm) exceeded 24,910 particles/L (97 μg/m3), positively correlating with the prevalence of cultivable bioaerosols. These bioaerosols contained cultivable bacterial species such as pathogenic Staphylococcus aureus, Staphylococcus epidermidis, Streptococcus pneumoniae, and Escherichia coli. Bioaerosol concentrations increased by 53.0% during 50-mm snow extremes. Taxonomic analysis revealed that Pseudomonas, Staphylococcus, and Veillonella were the most abundant bacterial taxa in the initial snowmelt samples during winter precipitation. However, their abundance decreased by 87.6% as snowing continued (24 h). Reduced water base cation concentration also led to a 1.15-fold increase in the Shannon index, indicating a similar yet heightened bacterial diversity. Seasonally, Pedobacter and Massilia showed higher relative abundance (25% and 18%, respectively), presenting increased bacterial transmission to the soil. Furthermore, Pseudomonas was identified in 60% of spring snowstorm samples, suggesting long-distance dispersal of pathogenic bacteria. When these atmospheric aerosol particles carrying biological entities (0.65-1.1 μm) penetrated human alveoli, the calculated hazard ratio was 0.55, which as observed in inhalation exposures. Consequently, this study underscores the risk of seasonal precipitation-enhanced ambient bacterial transmission.

Progress in methods for the detection of viable Escherichia coli

Escherichia coli (E. coli) is a prevalent enteric bacterium and a necessary organism to monitor for food safety and environmental purposes. Developing efficient and specific methods is critical for detecting and monitoring viable E. coli due to its high prevalence. Conventional culture methods are often laborious and time-consuming, and they offer limited capability in detecting potentially harmful viable but non-culturable E. coli in the tested sample, which highlights the need for improved approaches. Hence, there is a growing demand for accurate and sensitive methods to determine the presence of viable E. coli. This paper scrutinizes various methods for detecting viable E. coli, including culture-based methods, molecular methods that target DNAs and RNAs, bacteriophage-based methods, biosensors, and other emerging technologies. The review serves as a guide for researchers seeking additional methodological options and aiding in the development of rapid and precise assays. Moving forward, it is anticipated that methods for detecting E. coli will become more stable and robust, ultimately contributing significantly to the improvement of food safety and public health.

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.

Rapid Determination of Positive-Negative Bacterial Infection Based on Micro-Hyperspectral Technology

To meet the demand for rapid bacterial detection in clinical practice, this study proposed a joint determination model based on spectral database matching combined with a deep learning model for the determination of positive-negative bacterial infection in directly smeared urine samples. Based on a dataset of 8124 urine samples, a standard hyperspectral database of common bacteria and impurities was established. This database, combined with an automated single-target extraction, was used to perform spectral matching for single bacterial targets in directly smeared data. To address the multi-scale features and the need for the rapid analysis of directly smeared data, a multi-scale buffered convolutional neural network, MBNet, was introduced, which included three convolutional combination units and four buffer units to extract the spectral features of directly smeared data from different dimensions. The focus was on studying the differences in spectral features between positive and negative bacterial infection, as well as the temporal correlation between positive-negative determination and short-term cultivation. The experimental results demonstrate that the joint determination model achieved an accuracy of 97.29%, a Positive Predictive Value (PPV) of 97.17%, and a Negative Predictive Value (NPV) of 97.60% in the directly smeared urine dataset. This result outperformed the single MBNet model, indicating the effectiveness of the multi-scale buffered architecture for global and large-scale features of directly smeared data, as well as the high sensitivity of spectral database matching for single bacterial targets. The rapid determination solution of the whole process, which combines directly smeared sample preparation, joint determination model, and software analysis integration, can provide a preliminary report of bacterial infection within 10 min, and it is expected to become a powerful supplement to the existing technologies of rapid bacterial detection.

On the use of deep learning for phase recovery

Phase recovery (PR) refers to calculating the phase of the light field from its intensity measurements. As exemplified from quantitative phase imaging and coherent diffraction imaging to adaptive optics, PR is essential for reconstructing the refractive index distribution or topography of an object and correcting the aberration of an imaging system. In recent years, deep learning (DL), often implemented through deep neural networks, has provided unprecedented support for computational imaging, leading to more efficient solutions for various PR problems. In this review, we first briefly introduce conventional methods for PR. Then, we review how DL provides support for PR from the following three stages, namely, pre-processing, in-processing, and post-processing. We also review how DL is used in phase image processing. Finally, we summarize the work in DL for PR and provide an outlook on how to better use DL to improve the reliability and efficiency of PR. Furthermore, we present a live-updating resource ( https://github.com/kqwang/phase-recovery ) for readers to learn more about PR.

Particle Tracking and Micromixing Performance Characterization with a Mobile Device

Strategies to stir and mix reagents in microfluid devices have evolved concomitantly with advancements in manufacturing techniques and sensing. While there is a large array of reported designs to combine and homogenize liquids, most of the characterization has been focused on setups with two inlets and one outlet. While this configuration is helpful to directly evaluate the effects of features and parameters on the mixing degree, it does not portray the conditions for experiments that involve more than two substances required to be subsequently combined. In this work, we present a mixing characterization methodology based on particle tracking as an alternative to the most common approach to measure homogeneity using the standard deviation of pixel intensities from a grayscale image. The proposed algorithm is implemented on a free and open-source mobile application (MIQUOD) for Android devices, numerically tested on COMSOL Multiphysics, and experimentally tested on a bidimensional split and recombine micromixer and a three-dimensional micromixer with sinusoidal grooves for different Reynolds numbers and geometrical features for samples with fluids seeded with red, blue, and green microparticles. The application uses concentration field data and particle track data to evaluate up to eleven performance metrics. Furthermore, with the insights from the experimental and numerical data, a mixing index for particles (mp) is proposed to characterize mixing performance for scenarios with multiple input reagents.

Label-free deep learning-based species classification of bacteria imaged by phase-contrast microscopy

Reliable detection and classification of bacteria and other pathogens in the human body, animals, food, and water is crucial for improving and safeguarding public health. For instance, identifying the species and its antibiotic susceptibility is vital for effective bacterial infection treatment. Here we show that phase contrast time-lapse microscopy combined with deep learning is sufficient to classify four species of bacteria relevant to human health. The classification is performed on living bacteria and does not require fixation or staining, meaning that the bacterial species can be determined as the bacteria reproduce in a microfluidic device, enabling parallel determination of susceptibility to antibiotics. We assess the performance of convolutional neural networks and vision transformers, where the best model attained a class-average accuracy exceeding 98%. Our successful proof-of-principle results suggest that the methods should be challenged with data covering more species and clinically relevant isolates for future clinical use.

Artificial intelligence-enabled quantitative phase imaging methods for life sciences

Quantitative phase imaging, integrated with artificial intelligence, allows for the rapid and label-free investigation of the physiology and pathology of biological systems. This review presents the principles of various two-dimensional and three-dimensional label-free phase imaging techniques that exploit refractive index as an intrinsic optical imaging contrast. In particular, we discuss artificial intelligence-based analysis methodologies for biomedical studies including image enhancement, segmentation of cellular or subcellular structures, classification of types of biological samples and image translation to furnish subcellular and histochemical information from label-free phase images. We also discuss the advantages and challenges of artificial intelligence-enabled quantitative phase imaging analyses, summarize recent notable applications in the life sciences, and cover the potential of this field for basic and industrial research in the life sciences.

Automatically showing microbial growth kinetics with a high-performance microbial growth analyzer

It is difficult to show microbial growth kinetics online when they grow in complex matrices. We presented a novel strategy to address this challenge by developing a high-performance microbial growth analyzer (HPMGA), which employed a unique 32-channel capacitively coupled contactless conductivity detector as a sensing element and fixed with a CellStatz software. It was capable of online showing accurate and repeatable growth curves of well-dispersed and bad-dispersed microbes, whether they grew in homogeneous simple culture broth or heterogeneous complex matrices. Moreover, it could automatically report key growth kinetics parameters. In comparison to optical density (OD), plate counting and broth microdilution (BMD) methods, we demonstrated its practicability in five scenarios: 1) the illustration of the growth, growth rate, and acceleration curves of Escherichia coli (E. coli); 2) the antimicrobial susceptibility testing (AST) of Oxacillin against Staphylococcus aureus (S. aureus); 3) the determination of Ag nanoparticle toxicity on Providencia rettgeri (P. rettgeri); 4) the characterization of milk fermentation; and 5) the enumeration of viable pathogenic Vibrio in shrimp body. Results highlighted that the HPMGA method had the advantages of universality and effectivity. This technology would significantly facilitate the routine analysis of microbial growth in many fields (biology, medicine, clinic, life, food, environment, and ecology), paving an avenue for microbiologists to achieve research goals that have been inhibited for years due to a lack of practical analytical methods.

High-throughput microbial culturomics using automation and machine learning

Pure bacterial cultures remain essential for detailed experimental and mechanistic studies in microbiome research, and traditional methods to isolate individual bacteria from complex microbial ecosystems are labor-intensive, difficult-to-scale and lack phenotype-genotype integration. Here we describe an open-source high-throughput robotic strain isolation platform for the rapid generation of isolates on demand. We develop a machine learning approach that leverages colony morphology and genomic data to maximize the diversity of microbes isolated and enable targeted picking of specific genera. Application of this platform on fecal samples from 20 humans yields personalized gut microbiome biobanks totaling 26,997 isolates that represented >80% of all abundant taxa. Spatial analysis on >100,000 visually captured colonies reveals cogrowth patterns between Ruminococcaceae, Bacteroidaceae, Coriobacteriaceae and Bifidobacteriaceae families that suggest important microbial interactions. Comparative analysis of 1,197 high-quality genomes from these biobanks shows interesting intra- and interpersonal strain evolution, selection and horizontal gene transfer. This culturomics framework should empower new research efforts to systematize the collection and quantitative analysis of imaging-based phenotypes with high-resolution genomics data for many emerging microbiome studies.

AI-enabled biosensing for rapid pathogen detection: From liquid food to agricultural water

Rapid pathogen detection in food and agricultural water is essential for ensuring food safety and public health. However, complex and noisy environmental background matrices delay the identification of pathogens and require highly trained personnel. Here, we present an AI-biosensing framework for accelerated and automated pathogen detection in various water samples, from liquid food to agricultural water. A deep learning model was used to identify and quantify target bacteria based on their microscopic patterns generated by specific interactions with bacteriophages. The model was trained on augmented datasets to maximize data efficiency, using input images of selected bacterial species, and then fine-tuned on a mixed culture. Model inference was performed on real-world water samples containing environmental noises unseen during model training. Overall, our AI model trained solely on lab-cultured bacteria achieved rapid (< 5.5 h) prediction with 80-100% accuracy on the real-world water samples, demonstrating its ability to generalize to unseen data. Our study highlights the potential applications in microbial water quality monitoring during food and agricultural processes.

Rapid and stain-free quantification of viral plaque via lens-free holography and deep learning

A plaque assay-the gold-standard method for measuring the concentration of replication-competent lytic virions-requires staining and usually more than 48 h of runtime. Here we show that lens-free holographic imaging and deep learning can be combined to expedite and automate the assay. The compact imaging device captures phase information label-free at a rate of approximately 0.32 gigapixels per hour per well, covers an area of about 30 × 30 mm2 and a 10-fold larger dynamic range of virus concentration than standard assays, and quantifies the infected area and the number of plaque-forming units. For the vesicular stomatitis virus, the automated plaque assay detected the first cell-lysing events caused by viral replication as early as 5 h after incubation, and in less than 20 h it detected plaque-forming units at rates higher than 90% at 100% specificity. Furthermore, it reduced the incubation time of the herpes simplex virus type 1 by about 48 h and that of the encephalomyocarditis virus by about 20 h. The stain-free assay should be amenable for use in virology research, vaccine development and clinical diagnosis.

eXtreme gradient boosting-based classification of bacterial mixtures in water and milk using wireless microscopic imaging of quorum sensing peptide-conjugated particles

Numerous bacteria can cause water- and foodborne diseases and are often found in bacterial mixtures, making their detection challenging. Specific bioreceptors or selective growth media are necessary for most bacterial detection methods. In this work, we collectively used five quorum sensing-based peptides identified from bacterial biofilms to identify 10 different bacterial species (Bacillus subtilis, Campylobacter jejuni, Enterococcus faecium, Escherichia coli, Legionella pneumophila, Listeria monocytogenes, Pseudomonas aeruginosa, Salmonella Typhimurium, Staphylococcus aureus, Vibrio parahaemolyticus) and their mixtures in water and milk. Four different machine learning classification methods were used: k-nearest neighbors (k-NN), decision tree (DT), support vector machine (SVM), and eXtreme Gradient Boosting (XGBoost). Peptides were crosslinked to submicron particles, and peptide-bacteria interactions on paper microfluidic chips caused the particle aggregation. A wireless, pocket fluorescence microscope (interfaced with a smartphone) counted such particle aggregations. XGBoost showed the best accuracy of 83.75% in identifying bacterial species from water samples using 320 different datasets and 91.67% from milk samples using 140 different datasets (5 peptide features per dataset). Each peptide's contribution to correct classification was evaluated. The results were concentration-dependent, allowing the identification of a dominant species from bacterial mixtures. Using XGBoost and the previous milk database, we tested 14 blind samples of various bacterial mixtures in milk samples, with an accuracy of 81.55% to predict the dominant species. The entire process could be completed within a half hour. The demonstrated system can provide a handheld, low-cost, easy-to-operate tool for potential hygiene spot-checks, public health, or personal healthcare.

Lens Free Holographic Imaging for Urinary Tract Infection Screening

OBJECTIVE

The diagnosis of urinary tract infection (UTI) currently requires precise specimen collection, handling infectious human waste, controlled urine storage, and timely transportation to modern laboratory equipment for analysis. Here we investigate holographic lens free imaging (LFI) to show its promise for enabling automatic urine analysis at the patient bedside.

METHODS

We introduce an LFI system capable of resolving important urine clinical biomarkers such as red blood cells, white blood cells, crystals, and casts in 2 mm thick urine phantoms.

RESULTS

This approach is sensitive to the particulate concentrations relevant for detecting several clinical urine abnormalities such as hematuria and pyuria, linearly correlating to ground truth hemacytometer measurements with R 2 = 0.9941 and R 2 = 0.9973, respectively. We show that LFI can estimate E. coli concentrations of 10 3 to 10 5 cells/mL by counting individual cells, and is sensitive to concentrations of 10 5 cells/mL to 10 8 cells/mL by analyzing hologram texture. Further, LFI measurements of blood cell concentrations are relatively insensitive to changes in bacteria concentrations of over seven orders of magnitude. Lastly, LFI reveals clear differences between UTI-positive and UTI-negative urine from human patients.

CONCLUSION

LFI is sensitive to clinically-relevant concentrations of bacteria, blood cells, and other sediment in large urine volumes.

SIGNIFICANCE

Together, these results show promise for LFI as a tool for urine screening, potentially offering early, point-of-care detection of UTI and other pathological processes.

Highly sensitive quantitative phase microscopy and deep learning aided with whole genome sequencing for rapid detection of infection and antimicrobial resistance

Current state-of-the-art infection and antimicrobial resistance (AMR) diagnostics are based on culture-based methods with a detection time of 48-96 h. Therefore, it is essential to develop novel methods that can do real-time diagnoses. Here, we demonstrate that the complimentary use of label-free optical assay with whole-genome sequencing (WGS) can enable rapid diagnosis of infection and AMR. Our assay is based on microscopy methods exploiting label-free, highly sensitive quantitative phase microscopy (QPM) followed by deep convolutional neural networks-based classification. The workflow was benchmarked on 21 clinical isolates from four WHO priority pathogens that were antibiotic susceptibility tested, and their AMR profile was determined by WGS. The proposed optical assay was in good agreement with the WGS characterization. Accurate classification based on the gram staining (100% recall for gram-negative and 83.4% for gram-positive), species (98.6%), and resistant/susceptible type (96.4%), as well as at the individual strain level (100% sensitivity in predicting 19 out of the 21 strains, with an overall accuracy of 95.45%). The results from this initial proof-of-concept study demonstrate the potential of the QPM assay as a rapid and first-stage tool for species, strain-level classification, and the presence or absence of AMR, which WGS can follow up for confirmation. Overall, a combined workflow with QPM and WGS complemented with deep learning data analyses could, in the future, be transformative for detecting and identifying pathogens and characterization of the AMR profile and antibiotic susceptibility.

Future stem cell analysis: progress and challenges towards state-of-the art approaches in automated cells analysis

Background and Aims

A microscopic image has been used in cell analysis for cell type identification and classification, cell counting and cell size measurement. Most previous research works are tedious, including detailed understanding and time-consuming. The scientists and researchers are seeking modern and automatic cell analysis approaches in line with the current in-demand technology.

Objectives

This article provides a brief overview of a general cell and specific stem cell analysis approaches from the history of cell discovery up to the state-of-the-art approaches.

Methodology

A content description of the literature study has been surveyed from specific manuscript databases using three review methods: manuscript identification, screening, and inclusion. This review methodology is based on Prism guidelines in searching for originality and novelty in studies concerning cell analysis.

Results

By analysing generic cell and specific stem cell analysis approaches, current technology offers tremendous potential in assisting medical experts in performing cell analysis using a method that is less laborious, cost-effective, and reduces error rates.

Conclusion

This review uncovers potential research gaps concerning generic cell and specific stem cell analysis. Thus, it could be a reference for developing automated cells analysis approaches using current technology such as artificial intelligence and deep learning.

Spatio-temporal based deep learning for rapid detection and identification of bacterial colonies through lens-free microscopy time-lapses

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.

Biosensors and machine learning for enhanced detection, stratification, and classification of cells: a review

Biological cells, by definition, are the basic units which contain the fundamental molecules of life of which all living things are composed. Understanding how they function and differentiating cells from one another, therefore, is of paramount importance for disease diagnostics as well as therapeutics. Sensors focusing on the detection and stratification of cells have gained popularity as technological advancements have allowed for the miniaturization of various components inching us closer to Point-of-Care (POC) solutions with each passing day. Furthermore, Machine Learning has allowed for enhancement in the analytical capabilities of these various biosensing modalities, especially the challenging task of classification of cells into various categories using a data-driven approach rather than physics-driven. In this review, we provide an account of how Machine Learning has been applied explicitly to sensors that detect and classify cells. We also provide a comparison of how different sensing modalities and algorithms affect the classifier accuracy and the dataset size required.

Label-free 1D microfluidic dipstick counting of microbial colonies and bacteriophage plaques

Counting viable bacterial cells and functional bacteriophage is fundamental to microbiology underpinning research, surveillance, biopharmaceuticals and diagnostics. Colony forming unit (CFU) and plaque forming unit (PFU) counting still requires slow and laborious solid culture on agar in Petri dishes or plates. Here, we show that dip-stick microfluidic strips can be used without growth indicator dye for rapid and simple CFU ml^-1 and PFU ml^-1 measurement. We demonstrate for the first time that fluoropolymer microcapillaries combined with digital imaging allow bacteriophage plaques to be counted rapidly in a dip-and-test format. The microfluidic length scales offer a linear 1-dimensional alternative to a 2D solid agar medium surface, with colonies or plaques clearly visible as "dashes" or "gaps". An inexpensive open source darkfield biosensor system using Raspberry Pi imaging permits label-free detection and counting of colonies or plaques within 4-8 hours in a linear, liquid matrix within ∼200 μm inner diameter microcapillaries. We obtained full quantitative agreement between 1D microfluidic colony counting in dipsticks versus conventional 2D solid agar Petri dish plates for S. aureus and E. coli, and for T2 phage and phage K, but up to 6 times faster. Time-lapse darkfield imaging permitted detailed kinetic analysis of colony growth in the microcapillaries, providing new insight into microfluidic microbiology and colony growth, not possible with Petri dishes. Surprisingly, whilst E. coli colonies appeared earlier, subsequent colony expansion was faster along the microcapillaries for S. aureus. This may be explained by the microenvironment offered for 1D colony growth within microcapillaries, linked to a mass balance between nutrient (glucose) diffusion and bacterial growth kinetics. Counting individual colonies in liquid medium was not possible for motile strains that spread rapidly along the capillary, however inclusion of soft agar inhibited spreading, making this new simple dip-and-test counting method applicable to both motile and non-motile bacteria. Label-free dipstick colony and plaque counting has potential for many analytical microbial tasks, and the innovation of 1D colony counting has relevance to other microfluidic microbiology.

Simultaneous Optical Detection of Multiple Bacterial Species Using Nanometer-Scaled Metal-Organic Hybrids

This paper describes a simple strategy to identify bacteria using the optical properties of the nanohybrid structures (NHs) of polymer-coated metal nanoparticles (NPs). NHs, in which many small NPs are encapsulated in polyaniline particles, are useful optical labels because they produce strong scattered light. The light-scattering characteristics of NHs are strongly dependent on the constituent metal elements of NPs. Gold NHs (AuNHs), silver NHs (AgNHs), and copper NHs (CuNHs) produce white, reddish, and bluish scattered light, respectively. Moreover, unlike NPs, the color of the scattered light does not change even when NHs are aggregated. Introducing an antibody into NHs induces antigen-specific binding to cells, enabling the identification of bacteria based on light scattering. Multiple bacterial species adsorbed on the slide can be identified within a single field of view under a dark field microscope based on the color of the scattered light. Therefore, it is a useful development for safety risk assessments at manufacturing sites, such as those for foods, beverages, and drugs, and environmental surveys that require rapid detection of multiple bacteria.

A Deep-Learning Based System for Rapid Genus Identification of Pathogens under Hyperspectral Microscopic Images

Infectious diseases have always been a major threat to the survival of humanity. Additionally, they bring an enormous economic burden to society. The conventional methods for bacteria identification are expensive, time-consuming and laborious. Therefore, it is of great importance to automatically rapidly identify pathogenic bacteria in a short time. Here, we constructed an AI-assisted system for automating rapid bacteria genus identification, combining the hyperspectral microscopic technology and a deep-learning-based algorithm Buffer Net. After being trained and validated in the self-built dataset, which consists of 11 genera with over 130,000 hyperspectral images, the accuracy of the algorithm could achieve 94.9%, which outperformed 1D-CNN, 2D-CNN and 3D-ResNet. The AI-assisted system we developed has great potential in assisting clinicians in identifying pathogenic bacteria at the single-cell level with high accuracy in a cheap, rapid and automatic way. Since the AI-assisted system can identify the pathogenic genus rapidly (about 30 s per hyperspectral microscopic image) at the single-cell level, it can shorten the time or even eliminate the demand for cultivating. Additionally, the system is user-friendly for novices.

The Integration of Gold Nanoparticles with Polymerase Chain Reaction for Constructing Colorimetric Sensing Platforms for Detection of Health-Related DNA and Proteins

Polymerase chain reaction (PCR) is the standard tool in genetic information analysis, and the desirable detection merits of PCR have been extended to disease-related protein analysis. Recently, the combination of PCR and gold nanoparticles (AuNPs) to construct colorimetric sensing platforms has received considerable attention due to its high sensitivity, visual detection, capability for on-site detection, and low cost. However, it lacks a related review to summarize and discuss the advances in this area. This perspective gives an overview of established methods based on the combination of PCR and AuNPs for the visual detection of health-related DNA and proteins. Moreover, this work also addresses the future trends and perspectives for PCR-AuNP hybrid biosensors.

Technique Evolutions for Microorganism Detection in Complex Samples: A Review

Rapid detection of microorganisms is a major challenge in the medical and industrial sectors. In a pharmaceutical laboratory, contamination of medical products may lead to severe health risks for patients, such as sepsis. In the specific case of advanced therapy medicinal products, contamination must be detected as early as possible to avoid late production stop and unnecessary costs. Unfortunately, the conventional methods used to detect microorganisms are based on time-consuming and labor-intensive approaches. Therefore, it is important to find new tools to detect microorganisms in a shorter time frame. This review sums up the current methods and represents the evolution in techniques for microorganism detection. First, there is a focus on promising ligands, such as aptamers and antimicrobial peptides, cheaper to produce and with a broader spectrum of detection. Then, we describe methods achieving low limits of detection, thanks to Raman spectroscopy or precise handling of samples through microfluids devices. The last part is dedicated to techniques in real-time, such as surface plasmon resonance, preventing the risk of contamination. Detection of pathogens in complex biological fluids remains a scientific challenge, and this review points toward important areas for future research.

Generation of microbial colonies dataset with deep learning style transfer

We introduce an effective strategy to generate an annotated synthetic dataset of microbiological images of Petri dishes that can be used to train deep learning models in a fully supervised fashion. The developed generator employs traditional computer vision algorithms together with a neural style transfer method for data augmentation. We show that the method is able to synthesize a dataset of realistic looking images that can be used to train a neural network model capable of localising, segmenting, and classifying five different microbial species. Our method requires significantly fewer resources to obtain a useful dataset than collecting and labeling a whole large set of real images with annotations. We show that starting with only 100 real images, we can generate data to train a detector that achieves comparable results (detection mAP \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$=0.416$$\end{document}=0.416, and counting MAE \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$=4.49$$\end{document}=4.49) to the same detector but trained on a real, several dozen times bigger dataset (mAP \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$=0.520$$\end{document}=0.520, MAE \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$=4.31$$\end{document}=4.31), containing over 7 k images. We prove the usefulness of the method in microbe detection and segmentation, but we expect that it is general and flexible and can also be applicable in other domains of science and industry to detect various objects.