Search

Weibing Xu,1 Yuxin Di,1 Shengjing Chu,1 Zixuan Wang,1 Haitao Long,1 Lumei Pu,1 Runtian Ma,1 Yanwei Wang2 1College of Science, Gansu Agricultural University, Lanzhou, 730000, People’s Republic of China; 2Department of Chemical and Materials Engineering, School of Engineering and Digital Sciences, Nazarbayev University, Astana, 010000, KazakhstanCorrespondence: Weibing Xu, College of Science, Gansu Agricultural University, No. 1 Yingmen Village, Anning District, Lanzhou, 730000, People’s Republic of China, Tel/Fax +86 931 7631212, Email xuwb@gsau.edu.cnBackground: Combination therapy employing multiple drugs has been shown to enhance the efficacy of cancer treatment. Chlorambucil (Chl) and 6-mercaptopurine (6MP) are the first-line medicines for chronic lymphocytic leukemia and ovarian cancer. However, both were limited by their short half-life of disintegration, unsatisfactory water solubility, and adverse reactions.Methods: In this work, the drug Chl and 6MP were introduced into the polymerized N-(2-hydroxypropyl) methacrylamide (polyHPMA) by pH and glutathione responsive linker to construct the polymer nanodrug delivery system for effective co-delivery.Results: The drug load capacities, release, morphology, and cytotoxicity of the pro-drug were systematic. The two drugs showed satisfactory synergism with a combination index of 0.81, and a better ability to induce apoptosis. In and ex vivo fluorescence imaging showed a rapid systemic distribution of the conjugate within mice, majorly metabolized by liver and kidneys and eliminated after 24 hr. No significant pathological damage was observed in the major organs. This polymeric prodrug system holds promise for improved therapeutic efficiency and reduced side effects through the synergistic delivery of various chemotherapeutics.Conclusion: The introduction of HPMA as a carrier not only enhanced the solubility and biocompatibilities of Chl and 6 MP but also improved their drug effect. This strategy might be a promising alternative for constructing multi-drug-release system. Keywords: polymer prodrug, co-delivery, chlorambucil, mercaptopurine, fluorescence images, cancer treatment

Aflatoxin contamination in cacao is a significant problem in terms of trade losses and health effects. This calls for the need for a non-invasive, precise, and effective detection strategy. This research contribution is to determine the best deep-learning model to classify the aflatoxin contamination level in cocoa beans based on fluorescence images and deep learning to improve performance in the classification. The process involved inoculating and incubating Aspergillus flavus (6mL/100g) to obtain aflatoxin-contaminated cocoa beans for 7 days during the incubation period. Liquid Mass Chromatography (LCMS) was used to quantify the aflatoxin in order to categorize the images into different levels including "free of aflatoxin", "contaminated below the limit", and "contaminated above the limit". 300 images were acquired through a mini studio equipped with UV lamps. The aflatoxin level was classified using several pre-trained CNN approaches which has high accuracy such as GoogLeNet, SqueezeNet, AlexNet, and ResNet50. The sensitivity analysis showed that the highest classification accuracy was found in the GoogLeNet model with optimizer: Adam and learning rate: 0.0001 by 96.42%. The model was tested using a testing dataset and obtain accuracy of 96% based on the confusion matrix. The findings indicate that combining CNN with fluorescence images improved the ability to classify the amount of aflatoxin contamination in cacao beans. This method has the potential to be more accurate and economical than the current approach, which could be adapted to reduce aflatoxin's negative effects on food safety and cacao trade losses. [ABSTRACT FROM AUTHOR]

Myelin is an essential component of the nervous system and myelin damage causes demyelination diseases. Myelin is a sheet of oligodendrocyte membrane wrapped around the neuronal axon. In the fluorescent images, experts manually identify myelin by co-localization of oligodendrocyte and axonal membranes that fit certain shape and size criteria. Because myelin wriggles along x-y-z axes, machine learning is ideal for its segmentation. However, machine-learning methods, especially convolutional neural networks (CNNs), require a high number of annotated images, which necessitate expert labor. To facilitate myelin annotation, we developed a workflow and software for myelin ground truth extraction from multi-spectral fluorescent images. Additionally, to the best of our knowledge, for the first time, a set of annotated myelin ground truths for machine learning applications were shared with the community.

Sweet peppers are popular worldwide due to their nutrition and taste. Conventional vegetable tracing methods have been trialed, but the application of such labels or tags can be laborious and expensive, making their commercial application impractical. What is needed is a label-free method that can identify features unique to each individual fruit. Our research team has noted that sweet peppers have unique textural fluorescence features when observed under UV light that could potentially be used as a label-free signature for identification of individual fruit as it travels through the postharvest supply chain. The objective of this research was to assess the feature of these sweet pepper features for identification purposes. The macroscopic and microscopic images were taken to characterize the fluorescence. The results indicate that all sweet peppers possess dot-like fluorescence features on their surface. Furthermore, it was observed that 93.60% of these features exhibited changes in fluorescence intensity within the cuticle layer during the growth of a pepper. These features on the macro-image are visible under 365 nm UV light, but challenging to be seen under white LEDs and to be classified from the fluorescence spectrum under 365 nm light. This research reported the fluorescence feature on the sweet pepper, which is invisible under white light. The results show that the uniqueness of fluorescent features on the surface of sweet peppers has the potential to become a traceability technology due to the presence of its unique physical modality. [ABSTRACT FROM AUTHOR]

Cell instance segmentation is a fundamental task for many biological applications, especially for packed cells in three-dimensional (3D) microscope images that can fully display cellular morphology. Image processing algorithms based on neural networks and feature engineering have enabled great progress in two-dimensional (2D) instance segmentation. However, current methods cannot achieve high segmentation accuracy for irregular cells in 3D images. In this study, we introduce a universal, morphology-based 3D instance segmentation algorithm called Crop Once Merge Twice (C1M2), which can segment cells from a wide range of image types and does not require nucleus images. C1M2 can be extended to quantify the fluorescence intensity of fluorescent proteins and antibodies and automatically annotate their expression levels in individual cells. Our results suggest that C1M2 can serve as a tissue cytometry for 3D histopathological assays by quantifying fluorescence intensity with spatial localization and morphological information. [ABSTRACT FROM AUTHOR]

Cell counting in fluorescence microscopy is an essential task in biomedical research for analyzing cellular dynamics and studying disease progression. Traditional methods for cell counting involve manual counting or threshold-based segmentation, which are time-consuming and prone to human error. Recently, deep learning-based object detection methods have shown promising results in automating cell counting tasks. However, the existing methods mainly focus on segmentation-based techniques that require a large amount of labeled data and extensive computational resources. In this paper, we propose a novel approach to detect and count multiple-size cells in a fluorescence image slide using You Only Look Once version 5 (YOLOv5) with a feature pyramid network (FPN). Our proposed method can efficiently detect multiple cells with different sizes in a single image, eliminating the need for pixel-level segmentation. We show that our method outperforms state-of-the-art segmentation-based approaches in terms of accuracy and computational efficiency. The experimental results on publicly available datasets demonstrate that our proposed approach achieves an average precision of 0.8 and a processing time of 43.9 ms per image. Our approach addresses the research gap in the literature by providing a more efficient and accurate method for cell counting in fluorescence microscopy that requires less computational resources and labeled data. [ABSTRACT FROM AUTHOR]

Myelin is an essential component of the nervous system and myelin damage causes demyelination diseases. Myelin is a sheet of oligodendrocyte membrane wrapped around the neuronal axon. In the fluorescent images, experts manually identify myelin by co-localization of oligodendrocyte and axonal membranes that fit certain shape and size criteria. Because myelin wriggles along x-y-z axes, machine learning is ideal for its segmentation. However, machine-learning methods, especially convolutional neural networks (CNNs), require a high number of annotated images, which necessitate expert labor. To facilitate myelin annotation, we developed a workflow and software for myelin ground truth extraction from multi-spectral fluorescent images. Additionally, to the best of our knowledge, for the first time, a set of annotated myelin ground truths for machine learning applications were shared with the community.

Myelin is an essential component of the nervous system and myelin damage causes demyelination diseases. Myelin is a sheet of oligodendrocyte membrane wrapped around the neuronal axon. In the fluorescent images, experts manually identify myelin by co-localization of oligodendrocyte and axonal membranes that fit certain shape and size criteria. Because myelin wriggles along x-y-z axes, machine learning is ideal for its segmentation. However, machine-learning methods, especially convolutional neural networks (CNNs), require a high number of annotated images, which necessitate expert labor. To facilitate myelin annotation, we developed a workflow and software for myelin ground truth extraction from multi-spectral fluorescent images. Additionally, to the best of our knowledge, for the first time, a set of annotated myelin ground truths for machine learning applications were shared with the community.

In this study, a novel method for adsorption and detection of nonylphenol (NP) was carried out using a molecularly imprinted polymers/carbon dots (MIP/CDs) coated on cotton fabrics (CF-MIP/CDs) and smartphone-based image analysis. The MIP/CDs using NP template molecule, 3-aminopropyltriethoxysilane (APTES) monomer and tetraethoxysilane (TEOS) cross-linker were simply prepared on the cotton fabrics. After removal of NP from the CF-MIP/CDs, the adsorption of NP depended on contact time and initial concentration. Compared with the corresponding non-imprinted polymers/carbon dots (CF-NIP/CDs), the CF-MIP/CDs exhibited higher adsorption capacity and selectivity toward NP. The pseudo-first-order and the Freundlich models provided the best description for NP adsorption. Due to high fluorescence emission, the CF-MIP/CDs were used as a sensor for highly selective NP detection. The fluorescence images were taken by smartphone and analyzed by ImageJ program for RGB measurement. The values of ΔGreen intensity were linearly proportional to NP concentration ranging from 100 to 1000 μg/L with limit of detection and quantification as 35.9 μg/L and 108.9 μg/L, respectively. The CF-MIP/CDs could keep stable in five weeks and use for 4 cycles. The proposed sensor could detect NP in soil samples with well recoveries (92–116%) and relative standard deviation (< 0.64%). [ABSTRACT FROM AUTHOR]

Abstract Background In cell biology, increasing focus has been directed to fast events at subcellular space with the advent of fluorescent probes. As an example, voltage sensitive dyes (VSD) have been used to measure membrane potentials. Yet, even the most recently developed genetically encoded voltage sensors have demanded exhausting signal averaging through repeated experiments to quantify action potentials (AP). This analysis may be further hampered in subcellular signals defined by small regions of interest (ROI), where signal-to-noise ratio (SNR) may fall substantially. Signal processing techniques like blind source separation (BSS) are designed to separate a multichannel mixture of signals into uncorrelated or independent sources, whose potential to separate ROI signal from noise has been poorly explored. Our aims are to develop a method capable of retrieving subcellular events with minimal a priori information from noisy cell fluorescence images and to provide it as a computational tool to be readily employed by the scientific community. Results In this paper, we have developed METROID (Morphological Extraction of Transmembrane potential from Regions Of Interest Device), a new computational tool to filter fluorescence signals from multiple ROIs, whose code and graphical interface are freely available. In this tool, we developed a new ROI definition procedure to automatically generate similar-area ROIs that follow cell shape. In addition, simulations and real data analysis were performed to recover AP and electroporation signals contaminated by noise by means of four types of BSS: Principal Component Analysis (PCA), Independent Component Analysis (ICA), and two versions with discrete wavelet transform (DWT). All these strategies allowed for signal extraction at low SNR (− 10 dB) without apparent signal distortion. Conclusions We demonstrate the great capability of our method to filter subcellular signals from noisy fluorescence images in a single trial, avoiding repeated experiments. We provide this novel biomedical application with a graphical user interface at https://doi.org/10.6084/m9.figshare.11344046.v1 , and its code and datasets are available in GitHub at https://github.com/zoccoler/metroid .

Myelin is an essential component of the nervous system and myelin damage causes demyelination diseases. Myelin is a sheet of oligodendrocyte membrane wrapped around the neuronal axon. In the fluorescent images, experts manually identify myelin by co-localization of oligodendrocyte and axonal membranes that fit certain shape and size criteria. Because myelin wriggles along x-y-z axes, machine learning is ideal for its segmentation. However, machine-learning methods, especially convolutional neural networks (CNNs), require a high number of annotated images, which necessitates expert labor. To facilitate myelin annotation, we developed a workflow and a software for myelin ground truth extraction from multi-spectral fluorescent images. Additionally, we shared a set of myelin ground truths annotated using this workflow.

Background: In cell biology, increasing focus has been directed to fast events at subcellular space with the advent of fluorescent probes. As an example, voltage sensitive dyes (VSD) have been used to measure membrane potentials. Yet, even the most recently developed genetically encoded voltage sensors have demanded exhausting signal averaging through repeated experiments to quantify action potentials (AP). This analysis may be further hampered in subcellular signals defined by small regions of interest (ROI), where signal-to-noise ratio (SNR) may fall substantially. Signal processing techniques like blind source separation (BSS) are designed to separate a multichannel mixture of signals into uncorrelated or independent sources, whose potential to separate ROI signal from noise has been poorly explored. Our aims are to develop a method capable of retrieving subcellular events with minimal a priori information from noisy cell fluorescence images and to provide it as a computational tool to be readily employed by the scientific community. Results: In this paper, we have developed METROID (Morphological Extraction of Transmembrane potential from Regions Of Interest Device), a new computational tool to filter fluorescence signals from multiple ROIs, whose code and graphical interface are freely available. In this tool, we developed a new ROI definition procedure to automatically generate similar-area ROIs that follow cell shape. In addition, simulations and real data analysis were performed to recover AP and electroporation signals contaminated by noise by means of four types of BSS: Principal Component Analysis (PCA), Independent Component Analysis (ICA), and two versions with discrete wavelet transform (DWT). All these strategies allowed for signal extraction at low SNR (− 10 dB) without apparent signal distortion. Conclusions: We demonstrate the great capability of our method to filter subcellular signals from noisy fluorescence images in a single trial, avoiding repeated experiments. We provide this novel biomedical application with a graphical user interface at 10.6084/m9.figshare.11344046.v1, and its code and datasets are available in GitHub at . [ABSTRACT FROM AUTHOR]

Clostridioides difficile infection (CDI) is an enteric bacterial disease that is increasing in incidence worldwide. Symptoms of CDI range from mild diarrhea to severe life-threatening inflammation of the colon. While antibiotics are standard-of-care treatments for CDI, they are also the biggest risk factor for development of CDI and recurrence. Therefore, novel therapies that successfully treat CDI and protect against recurrence are an unmet clinical need. Screening for novel drug leads is often tested by manual image analysis. The process is slow, tedious and is subject to human error and bias. So far, little work has focused on computer-aided screening for drug leads based on fluorescence images. Here, we propose a novel method to identify characteristic morphological changes in human fibroblast cells exposed to C. difficile toxins based on computer vision algorithms supported by deep learning methods. Classical image processing algorithms for the pre-processing stage are used together with an adjusted pre-trained deep convolutional neural network responsible for cell classification. In this study, we take advantage of transfer learning methodology by examining pre-trained VGG-19, ResNet50, Xception, and DenseNet121 convolutional neural network (CNN) models with adjusted, densely connected classifiers. We compare the obtained results with those of other machine learning algorithms and also visualize and interpret them. The proposed models have been evaluated on a dataset containing 369 images with 6112 cases. DenseNet121 achieved the highest results with a 93.5% accuracy, 92% sensitivity, and 95% specificity, respectively.

Context: Medical diagnosis and clinical decisions rely heavily on the histopathological evaluation of tissue samples, especially in oncology. Historically, classical histopathology has been the gold standard for tissue evaluation and assessment by pathologists. The most widely and commonly used dyes in histopathology are hematoxylin and eosin (H&E) as most malignancies diagnosis is largely based on this protocol. H&E staining has been used for more than a century to identify tissue characteristics and structures morphologies that are needed for tumor diagnosis. In many cases, as tissue is scarce in clinical studies, fluorescence imaging is necessary to allow staining of the same specimen with multiple biomarkers simultaneously. Since fluorescence imaging is a relatively new technology in the pathology landscape, histopathologists are not used to or trained in annotating or interpreting these images. Aims, Settings and Design: To allow pathologists to annotate these images without the need for additional training, we designed an algorithm for the conversion of fluorescence images to brightfield H&E images. Subjects and Methods: In this algorithm, we use fluorescent nuclei staining to reproduce the hematoxylin information and natural tissue autofluorescence to reproduce the eosin information avoiding the necessity to specifically stain the proteins or intracellular structures with an additional fluorescence stain. Statistical Analysis Used: Our method is based on optimizing a transform function from fluorescence to H&E images using least mean square optimization. Results: It results in high quality virtual H&E digital images that can easily and efficiently be analyzed by pathologists. We validated our results with pathologists by making them annotate tumor in real and virtual H&E whole slide images and we obtained promising results. Conclusions: Hence, we provide a solution that enables pathologists to assess tissue and annotate specific structures based on multiplexed fluorescence images. [ABSTRACT FROM AUTHOR]

• Unique features under UV light were captured. • Experiments were conducted to evaluate the proposed matching algorithm. • The success rate of single pepper tracing reached 81.3 %. • A label-free technology for tracing single green pepper in one package. In the food supply chain, vegetable traceability holds significant importance. Conventional methods for traceability rely on tag-based systems such as barcodes, QR codes, or RFID tags. However, these methods face challenges when it comes to tracing single vegetables, such as green peppers between the grading facility and greenhouse harvesting. Unlike fruits like melons and watermelons, green peppers lack visible unique features on their surface that can be used to identify individual vegetables. Through our research, we have discovered that the fluorescence images of green peppers display a unique texture on the surface, which provides the possibility of identifying single green pepper. We proposed a single pepper traceable method that combines pepper images from the greenhouse and postharvest stage under UV light using imaging features. Our experiments aimed to evaluate the method's performance, including feature description, tracing success rate, performance change with storage, and changes with different length of the green peppers. The results showed that the KAZE feature was suitable for describing the surface feature of a green pepper under UV light, achieving a feature-matching performance of 81.3 % success rate in tracking individual peppers in each of the 15 packages, each with 10 peppers, using images from the day after harvest and greenhouse images. Furthermore, the method's performance could be affected by the storage time and length of the peppers. The proposed method could be a cost-effective, accurate, and label-free method to achieve single green pepper level traceability in smart agriculture. [ABSTRACT FROM AUTHOR]

The major property of radiophotoluminescent glass dosimeter is its repeatable readout for the radiation measurements. Accordingly, the radiophotoluminescent glass dosimeter system will become one of the important dosimeters for dose measurement in the future. The aims of this study were to develop and characterize the self-fabricated radiophotoluminescent glass dosimeter reader system. The photomultiplier tube, charge-coupled device, nitrogen pulsed laser, band pass filter, 500 nm long pass filter and GD-302 glass dosimeter were used in this study. Based on our study, we found that the values of R-squared for both photomultiplier tube and charge-coupled device in reading system were 0.9968 and 0.9992, respectively, which shows that the relationship between luminescent signals or fluorescence images intensity of glass dosimeter and irradiation dose is direct proportion. These results suggest that the photomultiplier tube and charge-coupled device optical detector can be used in radiophotoluminescent glass dosimeter readout system. [ABSTRACT FROM AUTHOR]

Colloid transport in fractured rock formations is an important process impacting the fate of pollutants in the subsurface. Despite intensive and outstanding research on their transport phenomena, the impact of small-scale surface heterogeneity on colloid behavior at the fracture scale remains difficult to assess. In particular, there is relatively little direct experimental evidence on the impact of natural fracture surface heterogeneity on colloid transport. To investigate this, we developed an experimental setup allowing the direct visualization of fluorescent colloid transport in a flow cell containing a natural chalk rock sample while simultaneously monitoring effluent colloid concentrations. We used samples containing both a natural fracture surface and an artificially made smooth surface from the same chalk core. We characterized the roughness and chemical composition of both surface types and numerically calculated each surface's velocity field. From the experiments, we obtained direct images of colloid transport over the surfaces, from which we calculated their dispersion coefficients and quantified the residual deposition of colloids on the rock surface. We also measured the colloid breakthrough curves by collecting eluent samples from the flow cell outlet. The natural fracture surface exhibited larger physical and chemical heterogeneity than the smooth, artificially generated surface. The aperture variability across the natural surface led to preferential flow and colloid transport which was qualitatively apparent in the fluorescent images. The colloid transport patterns matched the calculated velocity fields well, directly linking the surface topography and aperture variation to colloid transport. Compared to the artificially made surface, the natural surface also showed higher dispersion coefficients, which corresponded to the colloids' earlier breakthrough from the flow cell. While we found differences between the elemental composition of the natural and artificially smooth surfaces, we could not observe their impact on the colloids' surface attachment and retention. The main novelty in this work is the coupling of direct colloid transport imaging, breakthrough curve measurements, and colloid surface deposition analyses, in a flow cell containing a natural carbonate rock sample. Our experimental setup can be used to further investigate the link between surface heterogeneity, both chemical and physical, and colloid transport and deposition in natural rock fractures. • Colloid transport experiments with heterogeneous and smooth rock fracture surfaces. • Fluorescence visualization of colloid transport related to their breakthrough curves. • Heterogeneous surface leads to preferential flow compared with the smooth surface. • Preferential flow leads to earlier breakthroughs and higher dispersion coefficients. • Dispersion coefficients show trends similar to experiments with synthetic fractures. [ABSTRACT FROM AUTHOR]

Against the increase of bacterial resistance to traditional antibiotics, antimicrobial peptides (AMP) are considered as promising alternatives. Bacterial biofilms are more resistant to antibiotics that their planktonic counterpart. The purpose of this study was to investigate the action of an AMP against a nascent bacterial biofilm. The activity of dermaseptin S4 derivative S4(1–16)M4Ka against 6 h-old Pseudomonas fluorescens biofilms was assessed by using a combination of Attenuated Total Reflectance–Fourier Transform InfraRed (ATR–FTIR) spectroscopy in situ and in real time, fluorescence microscopy using the Bac light™ kit, and Atomic Force Microscopy (AFM, imaging and force spectroscopy). After exposure to the peptide at three concentrations, different dramatic and fast changes over time were observed in the ATR–FTIR fingerprints reflecting a concentration-dependent action of the AMP. The ATR–FTIR spectra revealed major biochemical and physiological changes, adsorption/accumulation of the AMP on the bacteria, loss of membrane lipids, bacterial detachment, bacterial regrowth, or inhibition of biofilm growth. AFM allowed estimating at the nanoscale the effect of the AMP on the nanomechanical properties of the sessile bacteria. The bacterial membrane elasticity data measured by force spectroscopy were consistent with ATR–FTIR spectra, and they allowed suggesting a mechanism of action of this AMP on sessile P. fluorescens . The combination of these three techniques is a powerful tool for in situ and in real time monitoring the activity of AMPs against bacteria in a biofilm. [ABSTRACT FROM AUTHOR]

Lasers in space often suffer from light absorbing deposits forming inside the resonator. This effect is commonly called laser-induced contamination (LIC), and its mechanism can be compared with laser chemical vapor deposition. LIC experiments were carried out in a dedicated vacuum chamber using a q-switched 355-nm laser, hafnia- or silica-coated fused silica samples and contamination by epoxy adhesive outgassing, or toluene vapor in vacuum. The typical deposit formation was observed at different experimental conditions and analyzed by in situ laser-induced fluorescence imaging and ex situ white light interference microscopy. We determined the average growth rate during the first growth stage (bump-shaped growth) and analyzed it as a function of the laser fluence, sample nature, and used contamination. The data show that the band gap of the sample is important for the LIC process in the first growth stage. The light absorbed in the sample leads to a temperature rise that drives the deposit growth. This knowledge opens a new pathway to minimize LIC that is complementary to studies that aim to reduce the adsorption of contaminant molecules by making chemical surface treatments. [ABSTRACT FROM AUTHOR]

Abstract: The fluorescent quantum dot–polymer composites were fabricated by incorporating thioglycolic acid capped CdTe quantum dots into polyacrylamide via cross-linking agents. The CdTe–polyacrylamide composites were characterized by fluorescence spectrophotometer and fluorescence microscope. The result shows that the quantum dot–polymer composites show strong photoluminescence in aqueous solution. The photoluminescence spectrum of quantum dot–polymer composites exhibits a slight blue shift compared to that of initial CdTe quantum dots. The slight shift might be attributed to the covalently bonding between the carboxyl groups of thiolglycolic acid capped on CdTe quantum dots and the amide groups of the polyacrylamide chains. [Copyright &y& Elsevier]

Clostridioides difficile infection (CDI) is an enteric bacterial disease that is increasing in incidence worldwide. Symptoms of CDI range from mild diarrhea to severe life-threatening inflammation of the colon. While antibiotics are standard-of-care treatments for CDI, they are also the biggest risk factor for development of CDI and recurrence. Therefore, novel therapies that successfully treat CDI and protect against recurrence are an unmet clinical need. Screening for novel drug leads is often tested by manual image analysis. The process is slow, tedious and is subject to human error and bias. So far, little work has focused on computer-aided screening for drug leads based on fluorescence images. Here, we propose a novel method to identify characteristic morphological changes in human fibroblast cells exposed to C. difficile toxins based on computer vision algorithms supported by deep learning methods. Classical image processing algorithms for the pre-processing stage are used together with an adjusted pre-trained deep convolutional neural network responsible for cell classification. In this study, we take advantage of transfer learning methodology by examining pre-trained VGG-19, ResNet50, Xception, and DenseNet121 convolutional neural network (CNN) models with adjusted, densely connected classifiers. We compare the obtained results with those of other machine learning algorithms and also visualize and interpret them. The proposed models have been evaluated on a dataset containing 369 images with 6112 cases. DenseNet121 achieved the highest results with a 93.5% accuracy, 92% sensitivity, and 95% specificity, respectively. [ABSTRACT FROM AUTHOR]

The photodynamic impact of water-soluble zwitterionic zinc phthalocyanines (ZnPc1-4) was studied on pathogenic bacterial strains after specific light exposure (LED 665 nm). The structural differences between the studied ZnPc1-4 are in the positions and the numbers of substitution groups as well as in the bridging atoms (sulfur or oxygen) between substituents and macrocycle. The three peripherally substituted compounds (ZnPc1-3) are tetra-2-(N-propanesulfonic acid)oxypyridine (ZnPc1), tetra-2-(N-propanesulfonic acid)mercaptopyridine (ZnPc2), and octa-substituted 2-(N-propanesulfonic acid)mercaptopyridine (ZnPc3). The nonperipherally substituted compound is tetra-2-(N-propanesulfonic acid)mercaptopyridine (ZnPc4). The uptake and localization capability are studied on Gram (+) Enterococcus faecalis and Gram (-) Pseudomonas aeruginosa as suspensions and as 48 h biofilms. Relatively high accumulations of ZnPc1-4 show bacteria in suspensions with different cell density. The compounds have complete penetration in E. faecalis biofilms but with nonhomogenous distribution in P. aeruginosa biomass. The cytotoxicity test (Balb/c 3T3 Neutral Red Uptake) with ZnPc1-4 suggests the lack of dark toxicity on normal cells. However, only ZnPc3 has a minimal photocytotoxic effect toward Balb/c 3T3 cells and a comparable high potential in the photoinactivation of pathogenic bacterial species.

Context: Medical diagnosis and clinical decisions rely heavily on the histopathological evaluation of tissue samples, especially in oncology. Historically, classical histopathology has been the gold standard for tissue evaluation and assessment by pathologists. The most widely and commonly used dyes in histopathology are hematoxylin and eosin (H&E) as most malignancies diagnosis is largely based on this protocol. H&E staining has been used for more than a century to identify tissue characteristics and structures morphologies that are needed for tumor diagnosis. In many cases, as tissue is scarce in clinical studies, fluorescence imaging is necessary to allow staining of the same specimen with multiple biomarkers simultaneously. Since fluorescence imaging is a relatively new technology in the pathology landscape, histopathologists are not used to or trained in annotating or interpreting these images. Aims, Settings and Design: To allow pathologists to annotate these images without the need for additional training, we designed an algorithm for the conversion of fluorescence images to brightfield H&E images. Subjects and Methods: In this algorithm, we use fluorescent nuclei staining to reproduce the hematoxylin information and natural tissue autofluorescence to reproduce the eosin information avoiding the necessity to specifically stain the proteins or intracellular structures with an additional fluorescence stain. Statistical Analysis Used: Our method is based on optimizing a transform function from fluorescence to H&E images using least mean square optimization. Results: It results in high quality virtual H&E digital images that can easily and efficiently be analyzed by pathologists. We validated our results with pathologists by making them annotate tumor in real and virtual H&E whole slide images and we obtained promising results. Conclusions: Hence, we provide a solution that enables pathologists to assess tissue and annotate specific structures based on multiplexed fluorescence images. [ABSTRACT FROM AUTHOR]