Interfacing Cell-Based Assays in Environmental Scanning Electron Microscopy Using Dielectrophoresis

Comparing machine learning and deep learning regression frameworks for accurate prediction of dielectrophoretic force

An intelligent sensing framework using Machine Learning (ML) and Deep Learning (DL) architectures to precisely quantify dielectrophoretic force invoked on microparticles in a textile electrode-based DEP sensing device is reported. The prediction accuracy and generalization ability of the framework was validated using experimental results. Images of pearl chain alignment at varying input voltages were used to build deep regression models using modified ML and CNN architectures that can correlate pearl chain alignment patterns of Saccharomyces cerevisiae(yeast) cells and polystyrene microbeads to DEP force. Various ML models such as K-Nearest Neighbor, Support Vector Machine, Random Forest, Neural Networks, and Linear Regression along with DL models such as Convolutional Neural Network (CNN) architectures of AlexNet, ResNet-50, MobileNetV2, and GoogLeNet have been analyzed in order to build an effective regression framework to estimate the force induced on yeast cells and microbeads. The efficiencies of the models were evaluated using Mean Absolute Error, Mean Absolute Relative, Mean Squared Error, R-squared, and Root Mean Square Error (RMSE) as evaluation metrics. ResNet-50 with RMSPROP gave the best performance, with a validation RMSE of 0.0918 on yeast cells while AlexNet with ADAM optimizer gave the best performance, with a validation RMSE of 0.1745 on microbeads. This provides a baseline for further studies in the application of deep learning in DEP aided Lab-on-Chip devices.

Deep-Learning Based Estimation of Dielectrophoretic Force

The ability to accurately quantify dielectrophoretic (DEP) force is critical in the development of high-efficiency microfluidic systems. This is the first reported work that combines a textile electrode-based DEP sensing system with deep learning in order to estimate the DEP forces invoked on microparticles. We demonstrate how our deep learning model can process micrographs of pearl chains of polystyrene (PS) microbeads to estimate the DEP forces experienced. Numerous images obtained from our experiments at varying input voltages were preprocessed and used to train three deep convolutional neural networks, namely AlexNet, MobileNetV2, and VGG19. The performances of all the models was tested for their validation accuracies. Models were also tested with adversarial images to evaluate performance in terms of classification accuracy and resilience as a result of noise, image blur, and contrast changes. The results indicated that our method is robust under unfavorable real-world settings, demonstrating that it can be used for the direct estimation of dielectrophoretic force in point-of-care settings.

[Research progress in the application of external field separation technology and microfluidic technology in the separation of micro/nanoscales]

The micro/nanoscales concerns interactions of entities with sizes in the range of 0.1-100 μm, such as biological cells, proteins, and particles. The separation of micro/nanoscales has been of immense significance for drug development, early-stage cancer detection, and customized precision therapy. For example, in recent years, rapid advances in the field of cell therapy have necessitated the development of simple and effective cell separation techniques. The isolation technique allows the collection of the required stem cells from complex samples. With the development of materials science and precision medicine, the separation of particles is also critical. The key physicochemical properties of micro/nanoscales are highly dependent on their specific size, shape, functional group, and mobility (based on the charged characteristics), which control their performance in the separation system. The current demand has made the simultaneous innovation of a separation system and an on-line detection platform imperative. Accordingly, various analytical methods involving the use of external forces, such as the flow field, magnetic field, electric field, and acoustic field, have been used for micro/nanoscales separation. Based on the physical and chemical parameters of the separation materials, these analytical methods can select different external force fields for micro/nanoscales separation, enabling real-time, accurate, efficient, and selective separation. However, at present, most of the applied field separation technologies require complex equipment and a large sample amount. This makes it crucial to miniaturize and integrate separation technologies for low-cost, rapid, and accurate micro/nanoscales separation. Microfluidic technology is a representative micro/nanoscales separation technology. It requires only a small volume of liquid, making it cost-effective; its high throughput enables continuous separation and analysis; its fast response in a microchip can allow many reactions; and finally, the miniaturization of the device allows the coupling of multiple detectors with the microchip. With the continuous growth and progress of microfluidic technology, some microfluidic platforms are now able to achieve the non-destructive separation of cells. They also enable on-line detection, offer high separation efficiency, and allow rapid separation for different biological samples. This review primarily summarizes recent advances in microfluidic chips based on flow field, electric field, magnetic field, acoustic field, and field separation technologies to improve the micro/nanoscales separation efficiency. This review also discusses the various external force fields of micro/nanoscales, such as a microparticle, single cell separation of substances classified introduction, and summarizes the advantages and disadvantages of their application and development. Finally, the prospect of the combined application of external field separation technology and microfluidic technology in the early screening of cancer cells and for precise micro/nanoscales separation is discussed, and the advantages and potential applications of the combined technology are proposed.

In-situ preparation of plant samples in ESEM for energy dispersive x-ray microanalysis and repetitive observation in SEM and ESEM

The Extended Low Temperature Method (ELTM) for the in-situ preparation of plant samples in an environmental scanning electron microscope enables carrying out repetitive topographical and material analysis at a higher resolution in the vacuum conditions of a scanning electron microscope or in the low gas pressure conditions of an environmental scanning electron microscope. The method does not require any chemical intervention and is thus suitable for imaging delicate structures rarely observable with common treatment methods. The method enables both sample stabilization as close to their native state as possible, as well as the transfer of the same sample from a low vacuum to an atmospheric condition for sample storage or later study. It is impossible for wet samples in the environmental scanning electron microscope. Our studies illustrate the high applicability of the ELTM for different types of plant tissue, from imaging of plant waxes at higher resolution, the morphological study of highly susceptible early somatic embryos to the elemental microanalysis of root cells. The method established here provides a very fast, universal and inexpensive solution for plant sample treatment usable in a commercial environmental scanning electron microscope equipped with a cooling Peltier stage.

Advances in microfluidic devices made from thermoplastics used in cell biology and analyses

Silicon and glass were the main fabrication materials of microfluidic devices, however, plastics are on the rise in the past few years. Thermoplastic materials have recently been used to fabricate microfluidic platforms to perform experiments on cellular studies or environmental monitoring, with low cost disposable devices. This review describes the present state of the development and applications of microfluidic systems used in cell biology and analyses since the year 2000. Cultivation, separation/isolation, detection and analysis, and reaction studies are extensively discussed, considering only microorganisms (bacteria, yeast, fungi, zebra fish, etc.) and mammalian cell related studies in the microfluidic platforms. The advantages/disadvantages, fabrication methods, dimensions, and the purpose of creating the desired system are explained in detail. An important conclusion of this review is that these microfluidic platforms are still open for research and development, and solutions need to be found for each case separately.

Review: Microbial analysis in dielectrophoretic microfluidic systems

Infections caused by various known and emerging pathogenic microorganisms, including antibiotic-resistant strains, are a major threat to global health and well-being. This highlights the urgent need for detection systems for microbial identification, quantification and characterization towards assessing infections, prescribing therapies and understanding the dynamic cellular modifications. Current state-of-the-art microbial detection systems exhibit a trade-off between sensitivity and assay time, which could be alleviated by selective and label-free microbial capture onto the sensor surface from dilute samples. AC electrokinetic methods, such as dielectrophoresis, enable frequency-selective capture of viable microbial cells and spores due to polarization based on their distinguishing size, shape and sub-cellular compositional characteristics, for downstream coupling to various detection modalities. Following elucidation of the polarization mechanisms that distinguish bacterial cells from each other, as well as from mammalian cells, this review compares the microfluidic platforms for dielectrophoretic manipulation of microbials and their coupling to various detection modalities, including immuno-capture, impedance measurement, Raman spectroscopy and nucleic acid amplification methods, as well as for phenotypic assessment of microbial viability and antibiotic susceptibility. Based on the urgent need within point-of-care diagnostics towards reducing assay times and enhancing capture of the target organism, as well as the emerging interest in isolating intact microbials based on their phenotype and subcellular features, we envision widespread adoption of these label-free and selective electrokinetic techniques.

Dielectrophoretic applications for disease diagnostics using lab-on-a-chip platforms

Dielectrophoresis is a powerful technique used to distinguish distinct cellular identities in heterogeneous cell populations and to monitor changes in the cell state without the need for biochemical tags, including live and dead cells. Recent studies in the past decade have indicated that dielectrophoresis can be used to discriminate the disease state of cells by exploring the differences in the dielectric polarizabilities of the cells. Factors controlling the dielectric polarizability are dependent on the conductivity and permittivity of the cell and the suspending medium, the cell morphology, the internal structure, and the electric double layer effects associated with the charges on the cell surface. Diseased cells, such as those associated with malaria, cancer, dengue, anthrax and human African trypanosomiasis, could be spatially trapped by positive dielectrophoresis or spatially separated from other healthy cells by negative dielectrophoretic forces. The aim of this review was to provide a better and deeper understanding on how dielectrophoresis can be utilized to manipulate diseased cells. This review compiles and compares the significant findings obtained by researchers in manipulating abnormal or unhealthy cells.

Microfluidic device for a rapid immobilization of zebrafish larvae in environmental scanning electron microscopy

Small vertebrate model organisms have recently gained popularity as attractive experimental models that enhance our understanding of human tissue and organ development. Despite a large body of evidence using optical spectroscopy for the characterization of small model organism on chip-based devices, no attempts have been so far made to interface microfabricated technologies with environmental scanning electron microscopy (ESEM). Conventional scanning electron microscopy requires high vacuum environments and biological samples must be, therefore, submitted to many preparative procedures to dehydrate, fix, and subsequently stain the sample with gold-palladium deposition. This process is inherently low-throughput and can introduce many analytical artifacts. This work describes a proof-of-concept microfluidic chip-based system for immobilizing zebrafish larvae for ESEM imaging that is performed in a gaseous atmosphere, under low vacuum mode and without any need for sample staining protocols. The microfabricated technology provides a user-friendly and simple interface to perform ESEM imaging on zebrafish larvae. Presented lab-on-a-chip device was fabricated using a high-speed infrared laser micromachining in a biocompatible poly(methyl methacrylate) thermoplastic. It consisted of a reservoir with multiple semispherical microwells designed to hold the yolk of dechorionated zebrafish larvae. Immobilization of the larvae was achieved by a gentle suction generated during blotting of the medium. Trapping region allowed for multiple specimens to be conveniently positioned on the chip-based device within few minutes for ESEM imaging.

High resolution scanning electron microscopy of cells using dielectrophoresis

Ultrastructural analysis of cells can reveal valuable information about their morphological, physiological, and biochemical characteristics. Scanning electron microscopy (SEM) has been widely used to provide high-resolution images from the surface of biological samples. However, samples need to be dehydrated and coated with conductive materials for SEM imaging. Besides, immobilizing non-adherent cells during processing and analysis is challenging and requires complex fixation protocols. In this work, we developed a novel dielectrophoresis based microfluidic platform for interfacing non-adherent cells with high-resolution SEM at low vacuum mode. The system enables rapid immobilization and dehydration of samples without deposition of chemical residues over the cell surface. Moreover, it enables the on-chip chemical stimulation and fixation of immobilized cells with minimum dislodgement. These advantages were demonstrated for comparing the morphological changes of non-budding and budding yeast cells following Lyticase treatment.

Dielectrophoresis: applications and future outlook in point of care

Dielectrophoresis (DEP) is a label free, noninvasive, stand alone, rapid, and sensitive particle manipulation and characterization technique. Improvements in micro-electro-mechanical systems technology have enabled the biomedical applications of DEP over the past decades. By this way, integration of DEP into lab-on-a-chip systems has become achievable, creating a potential tool for point-of-care (POC) systems. DEP can be utilized in many different POC applications including early detection and prognosis of various cancer types, diagnosis of infectious diseases, blood cell analysis, and stem cell therapy. However, there are still some challenges to be resolved to have DEP-based devices available in POC market. Today, researchers have focused on these challenges to have this powerful theory as a solution for many POC applications. Here, DEP theory, cell modeling, and most common device structures are introduced briefly. Next, POC applications of DEP theory, such as cell (blood, cancer, stem, and fetal) and microorganism separation, manipulation, and enrichment for diagnosis and prognosis, are explained. Integration of DEP with other detection techniques to have more sensitive systems is summarized. Finally, future outlook for DEP-based systems are discussed with some challenges, which are currently preventing these systems to be a common tool for POC applications, and possible solutions.

Real-time cell viability assays using a new anthracycline derivative DRAQ7®

The exclusion of charged fluorescent dyes by intact cells has become a well-established assay for determining viability of cells. In search for a noninvasive fluorescent probe capable of long-term monitoring of cell death in real-time, we evaluated a new anthracycline derivative DRAQ7. The novel probe does not penetrate the plasma membrane of living cells but when the membrane integrity is compromised, it enters and binds readily to nuclear DNA to report cell death. It proved to be nontoxic to a panel of cancer cell lines grown continuously for up to 72 h and did not induce any detectable DNA damage signaling when analyzed using laser scanning microscopy and flow cytometry. The DRAQ7 provided a sensitive, real-time readout of cell death induced by a variety of stressors such as hypoxia, starvation, and drug-induced cytotoxicity. The overall responses to anticancer agents and resulting pharmacological dose-response profiles were not affected by the growth of tumor cells in the presence DRAQ7. Moreover, we for the first time introduced a near real-time microflow cytometric assay based on combination of DRAQ7 and mitochondrial inner membrane potential (ΔΨ(m) ) sensitive probe TMRM. We provide evidence that this low-dosage, real-time labeling procedure provides multiparameter and kinetic fingerprint of anticancer drug action.