Intracellular stress tomography reveals stress focusing and structural anisotropy in cytoskeleton of living cells

We describe a novel synchronous detection approach to map the transmission of mechanical stresses within the cytoplasm of an adherent cell. Using fluorescent protein-labeled mitochondria or cytoskeletal components as fiducial markers, we measured displacements and computed stresses in the cytoskeleton of a living cell plated on extracellular matrix molecules that arise in response to a small, external localized oscillatory load applied to transmembrane receptors on the apical cell surface. Induced synchronous displacements, stresses, and phase lags were found to be concentrated at sites quite remote from the localized load and were modulated by the preexisting tensile stress (prestress) in the cytoskeleton. Stresses applied at the apical surface also resulted in displacements of focal adhesion sites at the cell base. Cytoskeletal anisotropy was revealed by differential phase lags in X vs. Y directions. Displacements and stresses in the cytoskeleton of a cell plated on poly-L-lysine decayed quickly and were not concentrated at remote sites. These data indicate that mechanical forces are transferred across discrete cytoskeletal elements over long distances through the cytoplasm in the living adherent cell. mechanical forces; deformation; focal adhesion; microfilament     

Validation tool for traction force microscopy

Traction force microscopy (TFM) is commonly used to estimate cells' traction forces from the deformation that they cause on their substrate. The accuracy of TFM highly depends on the computational methods used to measure the deformation of the substrate and estimate the forces, and also on the specifics of the experimental set-up. Computer simulations can be used to evaluate the effect of both the computational methods and the experimental set-up without the need to perform numerous experiments. Here, we present one such TFM simulator that addresses several limitations of the existing ones. As a proof of principle, we recreate a TFM experimental set-up, and apply a classic 2D TFM algorithm to recover the forces. In summary, our simulator provides a valuable tool to study the performance, refine experimentally, and guide the extraction of biological conclusions from TFM experiments.     

Simulation and evaluation of 3D traction force microscopy

Measuring cell-generated forces by Traction Force Microscopy (TFM) has become a standard tool in cell mechanobiology. Although widely used in two dimensional (2D) experiments, only a few methods exist to measure traction in three-dimensional (3D) cell culture, since 3D volumetric high-resolution microscopy and more demanding computational approaches are required. Although it is commonly known that the selected experimental and computational setup highly influence the quality and accuracy of the results, no existing methods can adequately assess the errors involved in this process. We present a fully integrated simulation and evaluation platform that allows one to simulate TFM images and quantify errors of an applied approach for traction stress reconstruction, in order to improve experiments that attempt to measure mechanical interaction in cellular systems. In this context, we show that a careful parameter selection can decrease the reconstructed traction error by up to 40%.     

High Resolution,Large Deformation 3D Traction Force Microscopy

Traction Force Microscopy (TFM) is a powerful approach for quantifying cell-material interactions that over the last two decades has contributed significantly to our understanding of cellular mechanosensing and mechanotransduction. In addition, recent advances in three-dimensional (3D) imaging and traction force analysis (3D TFM) have highlighted the significance of the third dimension in influencing various cellular processes. Yet irrespective of dimensionality, almost all TFM approaches have relied on a linear elastic theory framework to calculate cell surface tractions. Here we present a new high resolution 3D TFM algorithm which utilizes a large deformation formulation to quantify cellular displacement fields with unprecedented resolution. The results feature some of the first experimental evidence that cells are indeed capable of exerting large material deformations, which require the formulation of a new theoretical TFM framework to accurately calculate the traction forces. Based on our previous 3D TFM technique, we reformulate our approach to accurately account for large material deformation and quantitatively contrast and compare both linear and large deformation frameworks as a function of the applied cell deformation. Particular attention is paid in estimating the accuracy penalty associated with utilizing a traditional linear elastic approach in the presence of large deformation gradients.     

Markerless 2D kinematic analysis of underwater running: A deep learning approach

Kinematic analysis is often performed with a camera system combined with reflective markers placed over bony landmarks. This method is restrictive (and often expensive), and limits the ability to perform analyses outside of the lab. In the present study, we used a markerless deep learning-based method to perform 2D kinematic analysis of deep water running, a task that poses several challenges to image processing methods. A single GoPro camera recorded sagittal plane lower limb motion. A deep neural network was trained using data from 17 individuals, and then used to predict the locations of markers that approximated joint centres. We found that 300–400 labelled images were sufficient to train the network to be able to position joint markers with an accuracy similar to that of a human labeler (mean difference < 3 pixels, around 1 cm). This level of accuracy is sufficient for many 2D applications, such as sports biomechanics, coaching/training, and rehabilitation. The method was sensitive enough to differentiate between closely-spaced running cadences (45–85 strides per minute in increments of 5). We also found high test–retest reliability of mean stride data, with between-session correlation coefficients of 0.90–0.97. Our approach represents a low-cost, adaptable solution for kinematic analysis, and could easily be modified for use in other movements and settings. Using additional cameras, this approach could also be used to perform 3D analyses. The method presented here may have broad applications in different fields, for example by enabling markerless motion analysis to be performed during rehabilitation, training or even competition environments.     

Continuum vs. spring network models of airway-parenchymal interdependence

The outward tethering forces exerted by the lung parenchyma on the airways embedded within it are potent modulators of the ability of the airway smooth muscle to shorten. Much of our understanding of these tethering forces is based on treating the parenchyma as an elastic continuum; yet, on a small enough scale, the lung parenchyma in two dimensions would seem to be more appropriately described as a discrete spring network. We therefore compared how the forces and displacements in the parenchyma surrounding a contracting airway are predicted to differ depending on whether the parenchyma is modeled as an elastic continuum or as a spring network. When the springs were arranged hexagonally to represent alveolar walls, the predicted parenchymal stresses and displacements propagated substantially farther away from the airway than when the springs were arranged in a triangular pattern or when the parenchyma was modeled as a continuum. Thus, to the extent that the parenchyma in vivo behaves as a hexagonal spring network, our results suggest that the range of interdependence forces due to airway contraction may have a greater influence than was previously thought.     

Prediction of interresidue contacts with DeepMetaPSICOV in CASP13

In this article, we describe our efforts in contact prediction in the CASP13 experiment. We employed a new deep learning-based contact prediction tool, DeepMetaPSICOV (or DMP for short), together with new methods and data sources for alignment generation. DMP evolved from MetaPSICOV and DeepCov and combines the input feature sets used by these methods as input to a deep, fully convolutional residual neural network. We also improved our method for multiple sequence alignment generation and included metagenomic sequences in the search. We discuss successes and failures of our approach and identify areas where further improvements may be possible. DMP is freely available at: https://github.com/psipred/DeepMetaPSICOV .     

Triflumuron induces cytotoxic effects on hepatic and renal human cell lines

Insect growth regulator insecticides are a new class of pesticides, commonly used around the world to control insect damages. Among those compounds, we focused our interest on triflumuron (TFM), which is less toxic than other conventional insecticides. However, not much is known about its toxic effects on mammalian systems. Therefore, our study aimed toward evaluating the cytotoxic and genotoxic effects of TFM using two different cell lines, the human renal embryonic cells (HEK 293) and hepatocytes (Hep G2). We showed, according to the 3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyltetrazolium bromide assay, that TFM reduced significantly the cell viability and increased the reactive oxygen species generation, malondialdehyde levels, and mitochondrial membrane potential in both cell lines. The antioxidant system was disturbed as assessed by the increased activities in both catalase and superoxide dismutase. We demonstrated also, that TFM is an inductor of DNA damages quantified by the comet assay. Moreover, we showed an overexpression of proapoptotic Bax and a decrease in antiapoptotic Bcl‐2 expression. As a conclusion, we demonstrate that the liver presents the major target organ to TFM, in which the cytotoxicity and the genotoxic effects were significantly higher in hepatic cells than in renal cells and by consequence its uses must be controlled.     

Improving cerebral microvascular image quality of optical coherence tomography angiography with deep learning-based segmentation

Optical coherence tomography angiography (OCTA) can map the microvascular networks of the cerebral cortices with micrometer resolution and millimeter penetration. However, the high scattering of the skull and the strong noise in the deep imaging region will distort the vasculature projections and decrease the OCTA image quality. Here, we proposed a deep learning-based segmentation method based on a U-Net convolutional neural network to extract the cortical region from the OCT image. The vascular networks were then visualized by three OCTA algorithms. The image quality of the vasculature projections was assessed by two metrics, including the peak signal-to-noise ratio (PSNR) and the contrast-to-noise ratio (CNR). The results show the accuracy of the cortical segmentation was 96.07%. The PSNR and CNR values increased significantly in the projections of the selected cortical regions. The OCTA incorporating the deep learning-based cortical segmentation can efficiently improve the image quality and enhance the vasculature clarity.     

Determining substrate displacement and cell traction fields--a new approach

This paper presents a new approach for the traction force microscopy (TFM) method which determines traction forces exerted by adherent cells on a thin, elastic polyacrylamide gel embedded with fluorescent microbeads. In this enhanced TFM method, a pattern recognition technique is first applied to match the pair of microbead embedded images before and after deformation, which subsequently provides the displacement field of the elastic substrate. Once the displacement field is obtained, the 3-D finite element method (FEM) is used to compute cell traction forces. The new TFM has been applied to determine traction forces of human tendon fibroblasts. Compared to existing TFM methods, the present method has the following advantages: (1) its displacement field obtained is associated with microbead movements; (2) it considers the finite thickness of the thin polyacrylamide gel and is therefore free from the infinite half-space approximation adopted by existing TFM methods; and (3) its computation procedure for determining cell traction forces is fast.     

Eigenstrain as a mechanical set-point of cells

Cell contraction regulates how cells sense their mechanical environment. We sought to identify the set-point of cell contraction, also referred to as tensional homeostasis. In this work, bovine aortic endothelial cells (BAECs), cultured on substrates with different stiffness, were characterized using traction force microscopy (TFM). Numerical models were developed to provide insights into the mechanics of cell–substrate interactions. Cell contraction was modeled as eigenstrain which could induce isometric cell contraction without external forces. The predicted traction stresses matched well with TFM measurements. Furthermore, our numerical model provided cell stress and displacement maps for inspecting the fundamental regulating mechanism of cell mechanosensing. We showed that cell spread area, traction force on a substrate, as well as the average stress of a cell were increased in response to a stiffer substrate. However, the cell average strain, which is cell type-specific, was kept at the same level regardless of the substrate stiffness. This indicated that the cell average strain is the tensional homeostasis that each type of cell tries to maintain. Furthermore, cell contraction in terms of eigenstrain was found to be the same for both BAECs and fibroblast cells in different mechanical environments. This implied a potential mechanical set-point across different cell types. Our results suggest that additional measurements of contractility might be useful for monitoring cell mechanosensing as well as dynamic remodeling of the extracellular matrix (ECM). This work could help to advance the understanding of the cell-ECM relationship, leading to better regenerative strategies.     

The biochemical composition of the actomyosin network sets the magnitude of cellular traction forces

The regulation of cellular force production relies on the complex interplay between a well-conserved set of proteins of the cytoskeleton: actin, myosin, and α-actinin. Despite our deep knowledge of the role of these proteins in force production at the molecular scale, our understanding of the biochemical regulation of the magnitude of traction forces generated at the entire-cell level has been limited, notably by the technical challenge of measuring traction forces and the endogenous biochemical composition in the same cell. In this study, we developed an alternative Traction-Force Microscopy (TFM) assay, which used a combination of hydrogel micropatterning to define cell adhesion and shape and an intermediate fixation/immunolabeling step to characterize strain energies and the endogenous protein contents in single epithelial cells. Our results demonstrated that both the signal intensity and the area of the Focal Adhesion (FA)–associated protein vinculin showed a strong positive correlation with strain energy in mature FAs. Individual contents from actin filament and phospho-myosin displayed broader deviation in their linear relationship to strain energies. Instead, our quantitative analyzes demonstrated that their relative amount exhibited an optimum ratio of phospho-myosin to actin, allowing maximum force production by cells. By contrast, although no correlation was identified between individual α-actinin content and strain energy, the ratio of α-actinin to actin filaments was inversely related to strain energy. Hence, our results suggest that, in the cellular model studied, traction-force magnitude is dictated by the relative numbers of molecular motors and cross-linkers per actin filament, rather than the amounts of an individual component in the cytoskeletal network. This assay offers new perspectives to study in more detail the complex interplay between the endogenous biochemical composition of individual cells and the force they produce.     

Deep learning tools and modeling to estimate the temporal expression of cell cycle proteins from 2D still images

Automatic characterization of fluorescent labeling in intact mammalian tissues remains a challenge due to the lack of quantifying techniques capable of segregating densely packed nuclei and intricate tissue patterns. Here, we describe a powerful deep learning-based approach that couples remarkably precise nuclear segmentation with quantitation of fluorescent labeling intensity within segmented nuclei, and then apply it to the analysis of cell cycle dependent protein concentration in mouse tissues using 2D fluorescent still images. First, several existing deep learning-based methods were evaluated to accurately segment nuclei using different imaging modalities with a small training dataset. Next, we developed a deep learning-based approach to identify and measure fluorescent labels within segmented nuclei, and created an ImageJ plugin to allow for efficient manual correction of nuclear segmentation and label identification. Lastly, using fluorescence intensity as a readout for protein concentration, a three-step global estimation method was applied to the characterization of the cell cycle dependent expression of E2F proteins in the developing mouse intestine.     

How Nucleus Mechanics and ECM Microstructure Influence the Invasion of Single Cells and Multicellular Aggregates

In order to move in a three-dimensional extracellular matrix, the nucleus of a cell must squeeze through the narrow spacing among the fibers and, by adhering to them, the cell needs to exert sufficiently strong traction forces. If the nucleus is too stiff, the spacing too narrow, or traction forces too weak, the cell is not able to penetrate the network. In this article, we formulate a mathematical model based on an energetic approach, for cells entering cylindrical channels composed of extracellular matrix fibers. Treating the nucleus as an elastic body covered by an elastic membrane, the energetic balance leads to the definition of a necessary criterion for cells to pass through the regular network of fibers, depending on the traction forces exerted by the cells (or possibly passive stresses), the stretchability of the nuclear membrane, the stiffness of the nucleus, and the ratio of the pore size within the extracellular matrix with respect to the nucleus diameter. The results obtained highlight the importance of the interplay between mechanical properties of the cell and microscopic geometric characteristics of the extracellular matrix and give an estimate for a critical value of the pore size that represents the physical limit of migration and can be used in tumor growth models to predict their invasive potential in thick regions of ECM.     

Semantic region of interest and species classification in the deep neural network feature domain

In this paper, we focus on animal object detection and species classification in camera-trap images collected in highly cluttered natural scenes. Using a deep neural network (DNN) model training for animal- background image classification, we analyze the input camera-trap images to generate a multi-level visual representation of the input image. We detect semantic regions of interest for animals from this representation using k-mean clustering and graph cut in the DNN feature domain. These animal regions are then classified into animal species using multi-class deep neural network model. According the experimental results, our method achieves 99.75% accuracy for classifying animals and background and 90.89% accuracy for classifying 26 animal species on the Snapshot Serengeti dataset, outperforming existing image classification methods.     

Automated analysis of high‐content microscopy data with deep learning

Existing computational pipelines for quantitative analysis of high‐content microscopy data rely on traditional machine learning approaches that fail to accurately classify more than a single dataset without substantial tuning and training, requiring extensive analysis. Here, we demonstrate that the application of deep learning to biological image data can overcome the pitfalls associated with conventional machine learning classifiers. Using a deep convolutional neural network (DeepLoc) to analyze yeast cell images, we show improved performance over traditional approaches in the automated classification of protein subcellular localization. We also demonstrate the ability of DeepLoc to classify highly divergent image sets, including images of pheromone‐arrested cells with abnormal cellular morphology, as well as images generated in different genetic backgrounds and in different laboratories. We offer an open‐source implementation that enables updating DeepLoc on new microscopy datasets. This study highlights deep learning as an important tool for the expedited analysis of high‐content microscopy data.     

Deep Convolution Neural Network Based System for Early Diagnosis of Alzheimer's Disease

ObjectivesAlzheimer's Disease (AD) is the most general type of dementia. In all leading countries, it is one of the primary reasons of death in senior citizens. Currently, it is diagnosed by calculating the MSME score and by the manual study of MRI Scan. Also, different machine learning methods are utilized for automatic diagnosis but existing has some limitations in terms of accuracy. So, main objective of this paper to include a preprocessing method before CNN model to increase the accuracy of classification.Materials and methodIn this paper, we present a deep learning-based approach for detection of Alzheimer's Disease from ADNI database of Alzheimer's disease patients, the dataset contains fMRI and PET images of Alzheimer's patients along with normal person's image. We have applied 3D to 2D conversion and resizing of images before applying VGG-16 architecture of Convolution neural network for feature extraction. Finally, for classification SVM, Linear Discriminate, K means clustering, and Decision tree classifiers are used.ResultsThe experimental result shows that the average accuracy of 99.95% is achieved for the classification of the fMRI dataset, while the average accuracy of 73.46% is achieved with the PET dataset. On comparing results on the basis of accuracy, specificity, sensitivity and on some other parameters we found that these results are better than existing methods.Conclusionsthis paper, suggested a unique way to increase the performance of CNN models by applying some preprocessing on image dataset before sending to CNN architecture for feature extraction. We applied this method on ADNI database and on comparing the accuracies with other similar approaches it shows better results.     

NF-RCNN: Heart localization and right ventricle wall motion abnormality detection in cardiac MRI

Convolutional neural networks (CNNs) are extensively used in cardiac image analysis. However, heart localization has become a prerequisite to these networks since it decreases the size of input images. Accordingly, recent CNNs benefit from deeper architectures in gaining abstract semantic information. In the present study, a deep learning-based method was developed for heart localization in cardiac MR images. Further, Network in Network (NIN) was used as the region proposal network (RPN) of the faster R-CNN, and then NIN Faster-RCNN (NF-RCNN) was proposed. NIN architecture is formed based on “MLPCONV” layer, a combination of convolutional network and multilayer perceptron (MLP). Therefore, it could deal with the complicated structures of MR images. Furthermore, two sets of cardiac MRI dataset were used to evaluate the network, and all the evaluation metrics indicated an absolute superiority of the proposed network over all related networks. In addition, FROC curve, precision-recall (PR) analysis, and mean localization error were employed to evaluate the proposed network. In brief, the results included an AUC value of 0.98 for FROC curve, a mean average precision of 0.96 for precision-recall curve, and a mean localization error of 6.17 mm. Moreover, a deep learning-based approach for the right ventricle wall motion analysis (WMA) was performed on the first dataset and the effect of the heart localization on this algorithm was studied. The results revealed that NF-RCNN increased the speed and decreased the required memory significantly.     

pyTFM: A tool for traction force and monolayer stress microscopy

Cellular force generation and force transmission are of fundamental importance for numerous biological processes and can be studied with the methods of Traction Force Microscopy (TFM) and Monolayer Stress Microscopy. Traction Force Microscopy and Monolayer Stress Microscopy solve the inverse problem of reconstructing cell-matrix tractions and inter- and intra-cellular stresses from the measured cell force-induced deformations of an adhesive substrate with known elasticity. Although several laboratories have developed software for Traction Force Microscopy and Monolayer Stress Microscopy computations, there is currently no software package available that allows non-expert users to perform a full evaluation of such experiments. Here we present pyTFM, a tool to perform Traction Force Microscopy and Monolayer Stress Microscopy on cell patches and cell layers grown in a 2-dimensional environment. pyTFM was optimized for ease-of-use; it is open-source and well documented (hosted at https://pytfm.readthedocs.io/) including usage examples and explanations of the theoretical background. pyTFM can be used as a standalone Python package or as an add-on to the image annotation tool ClickPoints. In combination with the ClickPoints environment, pyTFM allows the user to set all necessary analysis parameters, select regions of interest, examine the input data and intermediary results, and calculate a wide range of parameters describing forces, stresses, and their distribution. In this work, we also thoroughly analyze the accuracy and performance of the Traction Force Microscopy and Monolayer Stress Microscopy algorithms of pyTFM using synthetic and experimental data from epithelial cell patches.