DeepLeish: a deep learning based support system for the detection of Leishmaniasis parasite from Giemsa-stained microscope images.

BACKGROUND: Leishmaniasis is a vector-born neglected parasitic disease belonging to the genus Leishmania. Out of the 30 Leishmania species, 21 species cause human infection that affect the skin and the internal organs. Around, 700,000 to 1,000,000 of the newly infected cases and 26,000 to 65,000 deaths are reported worldwide annually. The disease exhibits three clinical presentations, namely, the cutaneous, muco-cutaneous and visceral Leishmaniasis which affects the skin, mucosal membrane and the internal organs, respectively. The relapsing behavior of the disease limits its diagnosis and treatment efficiency. The common diagnostic approaches follow subjective, error-prone, repetitive processes. Despite, an ever pressing need for an accurate detection of Leishmaniasis, the research conducted so far is scarce. In this regard, the main aim of the current research is to develop an artificial intelligence based detection tool for the Leishmaniasis from the Geimsa-stained microscopic images using deep learning method. METHODS: Stained microscopic images were acquired locally and labeled by experts. The images were augmented using different methods to prevent overfitting and improve the generalizability of the system. Fine-tuned Faster RCNN, SSD, and YOLOV5 models were used for object detection. Mean average precision (MAP), precision, and Recall were calculated to evaluate and compare the performance of the models. RESULTS: The fine-tuned YOLOV5 outperformed the other models such as Faster RCNN and SSD, with the MAP scores, of 73%, 54% and 57%, respectively. CONCLUSION: The currently developed YOLOV5 model can be tested in the clinics to assist the laboratorists in diagnosing Leishmaniasis from the microscopic images. Particularly, in low-resourced healthcare facilities, with fewer qualified medical professionals or hematologists, our AI support system can assist in reducing the diagnosing time, workload, and misdiagnosis. Furthermore, the dataset collected by us will be shared with other researchers who seek to improve upon the detection system of the parasite. The current model detects the parasites even in the presence of the monocyte cells, but sometimes, the accuracy decreases due to the differences in the sizes of the parasite cells alongside the blood cells. The incorporation of cascaded networks in future and the quantification of the parasite load, shall overcome the limitations of the currently developed system.

Evaluation of Novel Embroidered Textile-Electrodes Made from Hybrid Polyamide Conductive Threads for Surface EMG Sensing.

Recently, there has been an increase in the number of reports on textile-based dry electrodes that can detect biopotentials without the need for electrolytic gels. However, these textile electrodes have a higher electrode skin interface impedance due to the improper contact between the skin and the electrode, diminishing the reliability and repeatability of the sensor. To facilitate improved skin-electrode contact, the effects of load and holding contact pressure were monitored for an embroidered textile electrode composed of multifilament hybrid thread for its application as a surface electromyography (sEMG) sensor. The effect of the textile's inter-electrode distance and double layering of embroidery that increases the density of the conductive threads were studied. Electrodes embroidered onto an elastic strap were wrapped around the forearm with a hook and loop fastener and tested for their performance. Time domain features such as the Root Mean Square (RMS), Average Rectified Value (ARV), and Signal to Noise Ratio (SNR) were quantitatively monitored in relation to the contact pressure and load. Experiments were performed in triplicates, and the sEMG signal characteristics were observed for various loads (0, 2, 4, and 6 kg) and holding contact pressures (5, 10, and 20 mmHg). sEMG signals recorded with textile electrodes were comparable in amplitude to those recorded using typical Ag/AgCl electrodes (28.45 dB recorded), while the signal-to-noise ratios were, 11.77, 19.60, 19.91, and 20.93 dB for the different loads, and 21.33, 23.34, and 17.45 dB for different holding pressures. The signal quality increased as the elastic strap was tightened further, but a pressure higher than 20 mmHg is not recommended because of the discomfort experienced by the subjects during data collection.

Squamous Cell Carcinoma of Skin Cancer Margin Classification From Digital Histopathology Images Using Deep Learning.

OBJECTIVES: Now a days, squamous cell carcinoma (SCC) margin assessment is done by examining histopathology images and inspection of whole slide images (WSI) using a conventional microscope. This is time-consuming, tedious, and depends on experts' experience which may lead to misdiagnosis and mistreatment plans. This study aims to develop a system for the automatic diagnosis of skin cancer margin for squamous cell carcinoma from histopathology microscopic images by applying deep learning techniques. METHODS: The system was trained, validated, and tested using histopathology images of SCC cancer locally acquired from Jimma Medical Center Pathology Department from seven different skin sites using an Olympus digital microscope. All images were preprocessed and trained with transfer learning pre-trained models by fine-tuning the hyper-parameter of the selected models. RESULTS: The overall best training accuracy of the models become 95.3%, 97.1%, 89.8%, and 89.9% on EffecientNetB0, MobileNetv2, ResNet50, VGG16 respectively. In addition to this, the best validation accuracy of the models was 94.7%, 91.8%, 87.8%, and 86.7% respectively. The best testing accuracy of the models at the same epoch was 95.2%, 91.5%, 87%, and 85.5% respectively. From these models, EfficientNetB0 showed the best average training and testing accuracy than the other models. CONCLUSIONS: The system assists the pathologist during the margin assessment of SCC by decreasing the diagnosis time from an average of 25 minutes to less than a minute.

De-Speckling Breast Cancer Ultrasound Images Using a Rotationally Invariant Block Matching Based Non-Local Means (RIBM-NLM) Method.

The ultrasonic technique is an indispensable imaging modality for diagnosis of breast cancer in young women due to its ability in efficiently capturing the tissue properties, and decreasing nega-tive recognition rate thereby avoiding non-essential biopsies. Despite the advantages, ultrasound images are affected by speckle noise, generating fine-false structures that decrease the contrast of the images and diminish the actual boundaries of tissues on ultrasound image. Moreover, speckle noise negatively impacts the subsequent stages in image processing pipeline, such as edge detec-tion, segmentation, feature extraction, and classification. Previous studies have formulated vari-ous speckle reduction methods in ultrasound images; however, these methods suffer from being unable to retain finer edge details and require more processing time. In this study, we propose a breast ultrasound de-speckling method based on rotational invariant block matching non-local means (RIBM-NLM) filtering. The effectiveness of our method has been demonstrated by com-paring our results with three established de-speckling techniques, the switching bilateral filter (SBF), the non-local means filter (NLMF), and the optimized non-local means filter (ONLMF) on 250 images from public dataset and 6 images from private dataset. Evaluation metrics, including Self-Similarity Index Measure (SSIM), Peak Signal to Noise Ratio (PSNR), and Mean Square Error (MSE) were utilized to measure performance. With the proposed method, we were able to record average SSIM of 0.8915, PSNR of 65.97, MSE of 0.014, RMSE of 0.119, and computational speed of 82 seconds at noise variance of 20dB using the public dataset, all with p-value of less than 0.001 compared against NLMF, ONLMF, and SBF. Similarly, the proposed method achieved av-erage SSIM of 0.83, PSNR of 66.26, MSE of 0.015, RMSE of 0.124, and computational speed of 83 seconds at noise variance of 20dB using the private dataset, all with p-value of less than 0.001 compared against NLMF, ONLMF, and SBF.

Accurate Machine-Learning-Based classification of Leukemia from Blood Smear Images.

BACKGROUND: Conventional identification of blood disorders based on visual inspection of blood smears through microscope is time consuming, error-prone and is limited by hematologist's physical acuity. Therefore, an automated optical image processing system is required to support the clinical decision-making. MATERIALS AND METHODS: Blood smear slides (n = 250) were prepared from clinical samples, imaged and analyzed in Jimma Medical Center, Hematology department. Samples were collected, analyzed and preserved from out and in-patients. The system was able to categorize four common types of leukemia's such as acute and chronic myeloid leukemia; and acute and chronic lymphoblastic leukemia, through a robust image segmentation protocol, followed by classification using the support vector machine. RESULTS: The system was able to classify leukemia types with an accuracy, sensitivity, specificity of 97.69%, 97.86% and 100%, respectively for the test datasets, and 97.5%, 98.55% and 100%, respectively, for the validation datasets. In addition, the system also showed an accuracy of 94.75% for the WBC counts that include both lymphocytes and monocytes. The computer-assisted diagnosis system took less than one minute for processing and assigning the leukemia types, compared to an average period of 30 minutes by unassisted manual approaches. Moreover, the automated system complements the healthcare workers' in their efforts, by improving the accuracy rates in diagnosis from ∼70% to over 97%. CONCLUSION: Importantly, our module is designed to assist the healthcare facilities in the rural areas of sub-Saharan Africa, equipped with fewer experienced medical experts, especially in screening patients for blood associated diseases including leukemia.

The improved survival rate and cost-effectiveness of a 7-day continuous subcutaneous insulin infusion set.

AIMS: The purpose of this article is to compare the insulin cost-savings of the Medtronic Extended Infusion Set (or EIS, a.k.a. Extended Wear Infusion Set) designed and labeled for up to 7-day use with rapid-acting insulins to the current standard of care, 2- to 3-day infusion sets. METHODS: There are three major improvements (reducing insulin waste, plastic waste, and adverse events) with the extended duration of infusion set wear. This analysis focuses on cost savings from reduced insulin wastage during set changes. Studies published on insulin infusion set survival and EIS clinical trial data (NCT04113694) were used to estimate device lifetime performance using a Markov chain Monte Carlo model, including the assessment of adverse effects and device failure. Total costs associated with infusion set change or failure were systematically found in published literature or estimated based on physical usage, and the direct impact on insulin costs was calculated. RESULTS: Based on the model and clinical data, EIS users can expect to change their infusion sets about 75 fewer times than standard set users each year. The costs related to unrecoverable insulin during an infusion set and reservoir change in the US were estimated to range from $19.79 to $22.48, resulting in approximately $1324 to $1677 in annual cost-savings for the typical user from minimizing insulin wastage. LIMITATIONS: The study only assessed devices used within a monitored setting, that is, clinical trials. In addition, the variability associated with healthcare standards and costs and individual treatment variability including insulin dosages, contribute to the uncertainties with the calculations. CONCLUSIONS: Our analysis demonstrates that by extending the duration of infusion set wear, there may be substantial cost savings by reducing insulin wastage.

Quantitative analysis of blood cells from microscopic images using convolutional neural network.

Blood cell count provides relevant clinical information about different kinds of disorders. Any deviation in the number of blood cells implies the presence of infection, inflammation, edema, bleeding, and other blood-related issues. Current microscopic methods used for blood cell counting are very tedious and are highly prone to different sources of errors. Besides, these techniques do not provide full information related to blood cells like shape and size, which play important roles in the clinical investigation of serious blood-related diseases. In this paper, deep learning-based automatic classification and quantitative analysis of blood cells are proposed using the YOLOv2 model. The model was trained on 1560 images and 2703-labeled blood cells with different hyper-parameters. It was tested on 26 images containing 1454 red blood cells, 159 platelets, 3 basophils, 12 eosinophils, 24 lymphocytes, 13 monocytes, and 28 neutrophils. The network achieved detection and segmentation of blood cells with an average accuracy of 80.6% and a precision of 88.4%. Quantitative analysis of cells was done following classification, and mean accuracy of 92.96%, 91.96%, 88.736%, and 92.7% has been achieved in the measurement of area, aspect ratio, diameter, and counting of cells respectively.Graphical abstract Graphical abstract where the first picture shows the input image of blood cells seen under a compound light microscope. The second image shows the tools used like OpenCV to pre-process the image. The third image shows the convolutional neural network used to train and perform object detection. The 4th image shows the output of the network in the detection of blood cells. The last images indicate post-processing applied on the output image such as counting of each blood cells using the class label of each detection and quantification of morphological parameters like area, aspect ratio, and diameter of blood cells so that the final result provides the number of each blood cell types (seven) and morphological information providing valuable clinical information.

Corrigendum to "Isothermal Titration Calorimetry and Surface Plasmon Resonance Analysis Using the Dynamic Approach".

[This corrects the article DOI: 10.1016/j.bbrep.2019.100712.].

Corrigendum "Isothermal titration calorimetry and surface plasmon resonance analysis using the dynamic approach" [Biochem. Biophys. Rep. 21 (2020) 100712].

[This corrects the article DOI: 10.1016/j.bbrep.2019.100712.].

Isothermal titration calorimetry and surface plasmon resonance analysis using the dynamic approach.

Biophysical techniques such as isothermal titration calorimetry (ITC) and surface plasmon resonance (SPR) are routinely used to ascertain the global binding mechanisms of protein-protein or protein-ligand interaction. Recently, Dumas etal, have explicitly modelled the instrument response of the ligand dilution and analysed the ITC thermogram to obtain kinetic rate constants. Adopting a similar approach, we have integrated the dynamic instrument response with the binding mechanism to simulate the ITC profiles of equivalent and independent binding sites, equivalent and sequential binding sites and aggregating systems. The results were benchmarked against the standard commercial software Origin-ITC. Further, the experimental ITC chromatograms of 2'-CMP + RNASE and BH3I-1 + hBCLXL interactions were analysed and shown to be comparable with that of the conventional analysis. Dynamic approach was applied to simulate the SPR profiles of a two-state model, and could reproduce the experimental profile accurately.

Log D analysis using dynamic approach.

Log D the logarithm ( log10 ) of the distribution coefficient ( D ), is one of the important parameters used in Lipinski's rule to assess the druggability of a molecule in pharmaceutical formulations. The distribution of a molecule between a hydrophobic organic phase and an aqueous buffer phase is influenced by the pH of the buffer system. In this work, we used both the conventional algebraic method and the generalized 'dynamic' approach to model the distribution coefficient of amphoteric, diamino-monoprotic molecule and monoprotic acid in the presence of salt or co-solvent. We have shown the equivalence of these methods by analysing the recently reported experimental data of amphoteric molecules such as nalidixic acid, mebendazole, benazepril and telmisartan.

Liver injury-on-a-chip: microfluidic co-cultures with integrated biosensors for monitoring liver cell signaling during injury.

Tissue injury triggers complex communication between cells via secreted signaling molecules such as cytokines and growth factors. Discerning when and where these signals begin and how they propagate over time is very challenging with existing cell culture and analysis tools. The goal of this study was to develop new tools in the form of microfluidic co-cultures with integrated biosensors for local and continuous monitoring of secreted signals. Specifically, we focused on how alcohol injury affects TGF-ß signaling between two liver cell types, hepatocytes and stellate cells. Activation of stellate cells happens early during liver injury and is at the center of liver fibrosis. We demonstrated that alcohol injury to microfluidic co-cultures caused significantly higher levels of stellate cell activation compared to conditioned media and transwell injury experiments. This highlighted the advantage of the microfluidic co-culture: placement of two cell types in close proximity to ensure high local concentrations of injury-promoting secreted signals. Next, we developed a microsystem consisting of five chambers, two for co-culturing hepatocytes with stellate cells and three additional chambers containing miniature aptamer-modified electrodes for monitoring secreted TGF-ß. Importantly, the walls separating microfluidic chambers were actuatable; they could be raised or lowered to create different configurations of the device. The use of reconfigurable microfluidics and miniature biosensors revealed that alcohol injury causes hepatocytes to secrete TGF-ß molecules, which diffuse over to neighboring stellate cells and trigger production of additional TGF-ß from stellate cells. Our results lend credence to the emerging view of hepatocytes as active participants of liver injury. Broadly speaking, our microsystem makes it possible to monitor paracrine crosstalk between two cell types communicating via the same signaling molecule (e.g. TGF-ß).

Reconfigurable microfluidic device with integrated antibody arrays for capture, multiplexed stimulation, and cytokine profiling of human monocytes.

Monocytes represent a class of immune cells that play a key role in the innate and adaptive immune response against infections. One mechanism employed by monocytes for sensing foreign antigens is via toll-like receptors (TLRs)-transmembrane proteins that distinguish classes of foreign pathogens, for example, bacteria (TLR4, 5, and 9) vs. fungi (TLR2) vs. viruses (TLR3, 7, and 8). Binding of antigens activates a signaling cascade through TLR receptors that culminate in secretion of inflammatory cytokines. Detection of these cytokines can provide valuable clinical data for drug developers and disease investigations, but this usually requires a large sample volume and can be technically inefficient with traditional techniques such as flow cytometry, enzyme-linked immunosorbent assay, or luminex. This paper describes an approach whereby antibody arrays for capturing cells and secreted cytokines are encapsulated within a microfluidic device that can be reconfigured to operate in serial or parallel mode. In serial mode, the device represents one long channel that may be perfused with a small volume of minimally processed blood. Once monocytes are captured onto antibody spots imprinted into the floor of the device, the straight channel is reconfigured to form nine individually perfusable chambers. To prove this concept, the microfluidic platform was used to capture monocytes from minimally processed human blood in serial mode and then to stimulate monocytes with different TLR agonists in parallel mode. Three cytokines, tumor necrosis factor-α, interleukin (IL)-6, and IL-10, were detected using anti-cytokine antibody arrays integrated into each of the six chambers. We foresee further use of this device in applications such as pediatric immunology or drug/vaccine testing where it is important to balance small sample volume with the need for high information content.

A microsystem integrating photodegradable hydrogel microstructures and reconfigurable microfluidics for single-cell analysis and retrieval.

We developed a micropatterned photodegradable hydrogel array integrated with reconfigurable microfluidics to enable cell secretion analysis and cell retrieval at the single-cell level. The activity of protease molecules secreted from single cells was monitored using FRET peptides entrapped inside microfabricated compartments. Antibody-modified gel islands tethering cells to the surface could be degraded by UV exposure to release specific single cells of interest.

A mathematical method for extracting cell secretion rate from affinity biosensors continuously monitoring cell activity.

Our laboratory has previously developed miniature aptasensors that may be integrated at the site of a small group of cells for continuous detection of cell secreted molecules such as inflammatory cytokine interferon gamma (IFN-γ). In a system such as this, the signal measured at the sensor surfaces is a complex function of transport, reaction, as well as of cellular activity. Herein, we report on the development of a mathematical framework for extracting cell production rates from binding curves generated with affinity biosensors. This framework consisted of a diffusion-reaction model coupled to a root finding algorithm for determining cell production rates values causing convergence of a predetermined criterion. To experimentally validate model predictions, we deployed a microfluidic device with an integrated biosensor for measuring the IFN-γ release from CD4 T cells. We found close agreement between secretion rate observed theoretically and those observed experimentally. After taking into account the differences in sensor geometry and reaction kinetics, the method for cell secretion rate determination described in this paper may be broadly applied to any biosensor continuously measuring cellular activity.

Reconfigurable microfluidics with integrated aptasensors for monitoring intercellular communication.

We report the development of a microsystem integrating anti-TNF-α aptasensors with vacuum-actuatable microfluidic devices that may be used to monitor intercellular communications. Actuatable chambers were used to expose to mitogen a group of ~600 cells while not stimulating another group of monocytes only 600 µm away. Co-localizing groups of cells with miniature 300 µm diameter aptamer-modified electrodes enabled monitoring of TNF-α release from each group independently. The microsystem allowed observation of the sequence of events that included 1) mitogenic activation of the first group of monocytes to produce TNF-α, 2) diffusion of TNF-α to the location of the second group of cells and 3) activation of the second group of cells resulting in the production of TNF-α by these cells. Thus, we were able to experimentally verify reciprocal paracrine crosstalk between the two groups of cells secreting the same signalling molecule. Given the prevalence of such cellular communications during injury, cancer or immune response and the dearth of available monitoring techniques, the microsystem described here is envisioned to have significant impact on cell biology.

On-chip regeneration of aptasensors for monitoring cell secretion.

We report on the use of reconfigurable microfluidics for on-chip regeneration of aptasensors used for continuous monitoring of cell-secreted products.

Micropatterned sensing hydrogels integrated with reconfigurable microfluidics for detecting protease release from cells.

Matrix metalloproteinases (MMPs) play a central role in the breakdown of the extracellular matrix and are typically upregulated in cancer cells. The goal of the present study is to develop microwells suitable for capture of cells and detection of cell-secreted proteases. Hydrogel microwells comprised of poly(ethylene glycol) (PEG) were photopatterned on glass and modified with ligands to promote cell adhesion. To sense protease release, peptides cleavable by MMP9 were designed to contain a donor/acceptor FRET pair (FITC and DABCYL). These sensing molecules were incorporated into the walls of the hydrogel wells to enable a detection scheme where cells captured within the wells secreted protease molecules which diffused into the gel, cleaved the peptide, and caused a fluorescence signal to come on. By challenging sensing hydrogel microstructures to known concentrations of recombinant MMP9, the limit of detection was determined to be 0.625 nM with a linear range extending to 40 nM. To enhance sensitivity and to limit cross-talk between adjacent sensing sites, microwell arrays containing small groups (∼20 cells/well) of lymphoma cells were integrated into reconfigurable PDMS microfluidic devices. Using this combination of sensing hydrogel microwells and reconfigurable microfluidics, detection of MMP9 release from as few as 11 cells was demonstrated. Smart hydrogel microstructures capable of sequestering small groups of cells and sensing cell function have multiple applications ranging from diagnostics to cell/tissue engineering. Further development of this technology will include single-cell analysis and function-based cell sorting capabilities.

Micropatterned surfaces functionalized with electroactive peptides for detecting protease release from cells.

Matrix metalloproteinases (MMPs) regulate composition of the extracellular matrix and play a critical role in cancer, fibrosis, and wound healing. This article describes a novel peptide-based electrochemical biosensor for detecting activity of cell-secreted protease MMP9. In this sensing strategy, a peptide specific to MMP9 was modified with a redox label (methylene blue (MB)) and immobilized on microfabricated 300 µm diameter Au electrodes. Challenging the electrodes with known concentrations of MMP9 resulted in the cleavage of the MB containing peptide fragment and caused a decrease in electrical signal measured by square wave voltammetry (SWV). The limit of detection for MMP9 was determined to be 60 pM with a linear range extending to 50 nM. In preparation to detect cell-secreted MMP9, glass surfaces with Au electrode arrays were further micropatterned with poly(ethylene glycol) (PEG) gel to define annular cell adhesive regions next to electrodes and render the remainder of the surface nonfouling. The surfaces were further modified with CD14 antibody to promote attachment of monocytes. The peptide-modified electrode arrays were integrated into PDMS microfluidic devices and incubated with U-937 cells, transformed monocytes known to produce MMPs. These studies revealed a 3-fold higher electrochemical signal from ∼400 activated monocytes after 10 min activation compared to nonactivated monocytes. Whereas this article focuses on MMP9 detection, the general strategy of employing redox-labeled peptides on electrodes should be broadly applicable for detection of other proteases and should have clinical as well as basic science applications.