Efficiency-Accuracy Trade-Off in Light Field Estimation with Cost Volume Construction and Aggregation.

The Rich spatial and angular information in light field images enables accurate depth estimation, which is a crucial aspect of environmental perception. However, the abundance of light field information also leads to high computational costs and memory pressure. Typically, selectively pruning some light field information can significantly improve computational efficiency but at the expense of reduced depth estimation accuracy in the pruned model, especially in low-texture regions and occluded areas where angular diversity is reduced. In this study, we propose a lightweight disparity estimation model that balances speed and accuracy and enhances depth estimation accuracy in textureless regions. We combined cost matching methods based on absolute difference and correlation to construct cost volumes, improving both accuracy and robustness. Additionally, we developed a multi-scale disparity cost fusion architecture, employing 3D convolutions and a UNet-like structure to handle matching costs at different depth scales. This method effectively integrates information across scales, utilizing the UNet structure for efficient fusion and completion of cost volumes, thus yielding more precise depth maps. Extensive testing shows that our method achieves computational efficiency on par with the most efficient existing methods, yet with double the accuracy. Moreover, our approach achieves comparable accuracy to the current highest-accuracy methods but with an order of magnitude improvement in computational performance.

Toward Real-Time Animal Tracking with Integrated Stimulus Control for Automated Conditioning in Aquatic Eco-Neurotoxicology.

Aquatic eco-neurotoxicology is an emerging field that requires new analytical systems to study the effects of pollutants on animal behaviors. This is especially true if we are to gain insights into one of the least studied aspects: the potential perturbations that neurotoxicants can have on cognitive behaviors. The paucity of experimental data is partly caused by a lack of low-cost technologies for the analysis of higher-level neurological functions (e.g., associative learning) in small aquatic organisms. Here, we present a proof-of-concept prototype that utilizes a new real-time animal tracking software for on-the-fly video analysis and closed-loop, external hardware communications to deliver stimuli based on specific behaviors in aquatic organisms, spanning three animal phyla: chordates (fish, frog), platyhelminthes (flatworm), and arthropods (crustacean). The system's open-source software features an intuitive graphical user interface and advanced adaptive threshold-based image segmentation for precise animal detection. We demonstrate the precision of animal tracking across multiple aquatic species with varying modes of locomotion. The presented technology interfaces easily with low-cost and open-source hardware such as the Arduino microcontroller family for closed-loop stimuli control. The new system has potential future applications in eco-neurotoxicology, where it could enable new opportunities for cognitive research in diverse small aquatic model organisms.

Physically-based simulation of elastic-plastic fusion of 3D bioprinted spheroids.

Spheroids are microtissues containing cells organized in a spherical shape whose diameter is usually less than a millimetre. Depending on the properties of the environment they are placed in, some nearby spheroids spontaneously fuse and generate a tissue. Given their potential to mimic features typical of body parts and their ability to assemble by fusing in permissive hydrogels, they have been used as building blocks to 3D bioprint human tissue parts. Parameters controlling the shape and size of a bioprinted tissue using fusing spheroid cultures include cell composition, hydrogel properties, and their relative initial position. Hence, simulating, anticipating, and then controlling the spheroid fusion process is essential to control the shape and size of the bioprinted tissue. This study presents the first physically-based framework to simulate the fusion process of bioprinted spheroids. The simulation is based on elastic-plastic solid and fluid continuum mechanics models. Both models use the 'smoothed particle hydrodynamics' method, which is based on discretizing the continuous medium into a finite number of particles and solving the differential equations related to the physical properties (e.g. Navier-Stokes equation) using a smoothing kernel function. To further investigate the effects of such parameters on spheroid shape and geometry, we performed sensitivity and morphological analysis to validate our simulations within-vitrospheroids. Through ourin-silicosimulations by changing the aforementioned parameters, we show that the proposed models appropriately simulate the range of the elastic-plastic behaviours ofin-vitrofusing spheroids to generate tissues of desired shapes and sizes. Altogether, this study presented a physically-based simulation that can provide a framework for monitoring and controlling the geometrical shape of spheroids, directly impacting future research using spheroids for tissue bioprinting.

Light Field View Synthesis Using the Focal Stack and All-in-Focus Image.

Light field reconstruction and synthesis algorithms are essential for improving the lower spatial resolution for hand-held plenoptic cameras. Previous light field synthesis algorithms produce blurred regions around depth discontinuities, especially for stereo-based algorithms, where no information is available to fill the occluded areas in the light field image. In this paper, we propose a light field synthesis algorithm that uses the focal stack images and the all-in-focus image to synthesize a 9 × 9 sub-aperture view light field image. Our approach uses depth from defocus to estimate a depth map. Then, we use the depth map and the all-in-focus image to synthesize the sub-aperture views, and their corresponding depth maps by mimicking the apparent shifting of the central image according to the depth values. We handle the occluded regions in the synthesized sub-aperture views by filling them with the information recovered from the focal stack images. We also show that, if the depth levels in the image are known, we can synthesize a high-accuracy light field image with just five focal stack images. The accuracy of our approach is compared with three state-of-the-art algorithms: one non-learning and two CNN-based approaches, and the results show that our algorithm outperforms all three in terms of PSNR and SSIM metrics.

Glaucoma diagnosis using multi-feature analysis and a deep learning technique.

In this study, we aimed to facilitate the current diagnostic assessment of glaucoma by analyzing multiple features and introducing a new cross-sectional optic nerve head (ONH) feature from optical coherence tomography (OCT) images. The data (n = 100 for both glaucoma and control) were collected based on structural, functional, demographic and risk factors. The features were statistically analyzed, and the most significant four features were used to train machine learning (ML) algorithms. Two ML algorithms: deep learning (DL) and logistic regression (LR) were compared in terms of the classification accuracy for automated glaucoma detection. The performance of the ML models was evaluated on unseen test data, n = 55. An image segmentation pilot study was then performed on cross-sectional OCT scans. The ONH cup area was extracted, analyzed, and a new DL model was trained for glaucoma prediction. The DL model was estimated using five-fold cross-validation and compared with two pre-trained models. The DL model trained from the optimal features achieved significantly higher diagnostic performance (area under the receiver operating characteristic curve (AUC) 0.98 and accuracy of 97% on validation data and 96% on test data) compared to previous studies for automated glaucoma detection. The second DL model used in the pilot study also showed promising outcomes (AUC 0.99 and accuracy of 98.6%) to detect glaucoma compared to two pre-trained models. In combination, the result of the two studies strongly suggests the four features and the cross-sectional ONH cup area trained using deep learning have a great potential for use as an initial screening tool for glaucoma which will assist clinicians in making a precise decision.

Noise-Resilient Depth Estimation for Light Field Images Using Focal Stack and FFT Analysis.

Depth estimation for light field images is essential for applications such as light field image compression, reconstructing perspective views and 3D reconstruction. Previous depth map estimation approaches do not capture sharp transitions around object boundaries due to occlusions, making many of the current approaches unreliable at depth discontinuities. This is especially the case for light field images because the pixels do not exhibit photo-consistency in the presence of occlusions. In this paper, we propose an algorithm to estimate the depth map for light field images using depth from defocus. Our approach uses a small patch size of pixels in each focal stack image for comparing defocus cues, allowing the algorithm to generate sharper depth boundaries. Then, in contrast to existing approaches that use defocus cues for depth estimation, we use frequency domain analysis image similarity checking to generate the depth map. Processing in the frequency domain reduces the individual pixel errors that occur while directly comparing RGB images, making the algorithm more resilient to noise. The algorithm has been evaluated on both a synthetic image dataset and real-world images in the JPEG dataset. Experimental results demonstrate that our proposed algorithm outperforms state-of-the-art depth estimation techniques for light field images, particularly in case of noisy images.

Surface properties and the perception of color.

We examined whether perception of color saturation and lightness depends on the three-dimensional (3D) shape and surface gloss of surfaces rendered to have different hues. In Experiment 1, we parametrically varied specular roughness of predominantly planar surfaces with different mesoscopic relief heights. The orientation of surfaces was varied relative to the light source and observer. Observers matched perceived lightness and chroma (effectively saturation) using spherical objects rendered using CIE LCH color space. We observed strong interactions between perceived saturation and lightness with changes in surface orientation and surface properties (specular roughness and 3D relief height). Declines in saturation and increases in lightness were observed with increasing specular roughness. Changes in relief height had greater effects on perceived saturation and lightness for blue hues compared with reddish and greenish hues. Experiment 2 found inverse correlations between perceived gloss and specular roughness across conditions. Experiment 3 estimated perceived specular coverage and found that a weighted combination of perceived gloss and specular coverage could account for perceived color saturation and lightness, with different coefficients accounting for the perceptual experience for each of the three hue conditions. These findings suggest that perceived color saturation and lightness depend on the separation of specular highlights from diffuse shading informative of chromatic surface reflectance.

Single-cell Screening Method for the Selection and Recovery of Antibodies with Desired Specificities from Enriched Human Memory B Cell Populations.

The human antibody repertoire represents a largely untapped source of potential therapeutic antibodies and useful biomarkers. While current computational methods, such as next generation sequencing (NGS), yield enormous sets of data on the antibody repertoire at the sequence level, functional data is required to identify which sequences are relevant for a particular antigen or set of antigens. Here, we describe a method to identify and recover individual antigen-specific antibodies from peripheral blood mononuclear cells (PBMCs) from a human blood donor. This method utilizes an initial enrichment of mature B cells and requires a combination of phenotypic cell markers and fluorescently-labeled protein to isolate IgG memory B cells via flow cytometry. The heavy and light chain variable regions are then cloned and re-screened. Although limited to the memory B cell compartment, this method takes advantage of flow cytometry to interrogate millions of B cells and returns paired heavy and light chain sequences from a single cell in a format ready for expression and confirmation of specificity. Antibodies recovered with this method can be considered for therapeutic potential, but can also link specificity and function with bioinformatic approaches to assess the B cell repertoire within individuals.

Naturally occurring antibodies isolated from PD patients inhibit synuclein seeding in vitro and recognize Lewy pathology.

Deposition of α-synuclein into Lewy bodies and Lewy neurites is the hallmark of Parkinson's disease (PD). It is hypothesized that α-synuclein pathology spreads by a "prion-like" mechanism (i.e., by seeded aggregation or templated misfolding). Therefore, various extracellular α-synuclein conformers and/or posttranslational modifications may serve as biomarkers of disease or potential targets for novel interventions. To explore whether the antibody repertoires of PD patients contain anti-α-synuclein antibodies that can potentially be used as markers or immunotherapy, we interrogated peripheral IgG+ memory B cells from PD patients for reactivity to α-synuclein. In total, ten somatically mutated antibodies were recovered, suggesting the presence of an ongoing antigen-driven immune response. The three antibodies that had the highest affinity to recombinant full-length α-synuclein, aSyn-323.1, aSyn-336.1 and aSyn-338.1, were characterized further and shown to recognize epitopes in the C terminus of α-synuclein with binding affinities between 0.3 and 2.8 µM. Furthermore, all three antibodies were able to neutralize the "seeding" of intracellular synuclein aggregates in an in vitro α-synuclein seeding assay. Finally, differential reactivities were observed for all three human anti-α-synuclein antibodies across tissue treatment conditions by immunohistochemistry. Our results suggest that the memory B-cell repertoire of PD patients might represent a potential source of biomarkers and therapies.

Different STAT Transcription Complexes Drive Early and Delayed Responses to Type I IFNs.

IFNs, which transduce pivotal signals through Stat1 and Stat2, effectively suppress the replication of Legionella pneumophila in primary murine macrophages. Although the ability of IFN-γ to impede L. pneumophila growth is fully dependent on Stat1, IFN-αß unexpectedly suppresses L. pneumophila growth in both Stat1- and Stat2-deficient macrophages. New studies demonstrating that the robust response to IFN-αß is lost in Stat1-Stat2 double-knockout macrophages suggest that Stat1 and Stat2 are functionally redundant in their ability to direct an innate response toward L. pneumophila. Because the ability of IFN-αß to signal through Stat1-dependent complexes (i.e., Stat1-Stat1 and Stat1-Stat2 dimers) has been well characterized, the current studies focus on how Stat2 is able to direct a potent response to IFN-αß in the absence of Stat1. These studies reveal that IFN-αß is able to drive the formation of a Stat2 and IFN regulatory factor 9 complex that drives the expression of a subset of IFN-stimulated genes, but with substantially delayed kinetics. These observations raise the possibility that this pathway evolved in response to microbes that have devised strategies to subvert Stat1-dependent responses.

An iminosugar with potent inhibition of dengue virus infection in vivo.

The aim of the present study was to evaluate the ability of the iminosugar drug UV-4 to provide in vivo protection from lethal dengue virus (DENV) challenge. This study utilized a well-described model of dengue hemorrhagic fever/dengue shock syndrome (DHF/DSS)-like lethal disease in AG129 mice lacking the type I and II interferon receptors. Herein, we present UV-4 as a potent iminosugar for controlling DENV infection and disease in this mouse model. Specifically, administration of UV-4 reduced mortality, as well as viremia and viral RNA in key tissues, and cytokine storm. In addition, UV-4 treatment can be delayed, and it does not alter the anti-DENV antibody response. These results have set the foundation for development of UV-4 as a DENV-specific antiviral in phase I human clinical trials.

Tracking the evolution of dengue virus strains D2S10 and D2S20 by 454 pyrosequencing.

Dengue virus is the most prevalent mosquito-borne virus worldwide. In this study, we used pyrosequencing to analyze the whole viral genome of two mouse-adapted strains, D2S10 and D2S20, that induce a dengue hemorrhagic fever/dengue shock syndrome (DHF/DSS)-like lethal disease in mice lacking the type I and/or type II interferon receptors. Previous experiments with D2S10 indicated that N124D and K128E mutations in the envelope protein were responsible for the severe disease induced in mice compared to its parental strain PL046. Here we demonstrate that D2S20 is more virulent than D2S10 and captured the presence of five key amino acid mutations--T70I, N83D, and K122I in envelope (E), and A62T in nonstructural protein 2A (NS2A) and G605V in nonstructural protein 5 (NS5)--that may account for this. These findings set the foundation for further dissection of the viral determinants responsible for dengue disease manifestations in mouse models.

Inhibition of dengue virus infections in cell cultures and in AG129 mice by a small interfering RNA targeting a highly conserved sequence.

The dengue viruses (DENVs) exist as numerous genetic strains that are grouped into four antigenically distinct serotypes. DENV strains from each serotype can cause severe disease and threaten public health in tropical and subtropical regions worldwide. No licensed antiviral agent to treat DENV infections is currently available, and there is an acute need for the development of novel therapeutics. We found that a synthetic small interfering RNA (siRNA) (DC-3) targeting the highly conserved 5' cyclization sequence (5'CS) region of the DENV genome reduced, by more than 100-fold, the titers of representative strains from each DENV serotype in vitro. To determine if DC-3 siRNA could inhibit DENV in vivo, an "in vivo-ready" version of DC-3 was synthesized and tested against DENV-2 by using a mouse model of antibody-dependent enhancement of infection (ADE)-induced disease. Compared with the rapid weight loss and 5-day average survival time of the control groups, mice receiving the DC-3 siRNA had an average survival time of 15 days and showed little weight loss for approximately 12 days. DC-3-treated mice also contained significantly less virus than control groups in several tissues at various time points postinfection. These results suggest that exogenously introduced siRNA combined with the endogenous RNA interference processing machinery has the capacity to prevent severe dengue disease. Overall, the data indicate that DC-3 siRNA represents a useful research reagent and has potential as a novel approach to therapeutic intervention against the genetically diverse dengue viruses.

STAT2 mediates innate immunity to Dengue virus in the absence of STAT1 via the type I interferon receptor.

Dengue virus (DENV) is a mosquito-borne flavivirus, and symptoms of infection range from asymptomatic to the severe dengue hemorrhagic fever/dengue shock syndrome (DHF/DSS). High viral loads correlate with disease severity, and both type I & II interferons (IFNs) are crucial for controlling viral replication. We have previously reported that signal transducer and activator of transcription (STAT) 1-deficient mice are resistant to DENV-induced disease, but little is known about this STAT1-independent mechanism of protection. To determine the molecular basis of the STAT1-independent pathway, mice lacking STAT1, STAT2, or both STAT1 and STAT2 were infected with a virulent mouse-adapted strain of DENV2. In the first 72 hours of infection, the single-deficient mice lacking STAT1 or STAT2 possessed 50-100 fold higher levels of viral RNA than wild type mice in the serum, spleen, and other visceral tissues, but remained resistant to DENV-induced death. In contrast, the double-deficient mice exhibited the early death phenotype previously observed in type I and II IFN receptor knockout mice (AG129), indicating that STAT2 is the mediator of the STAT1-independent host defense mechanism. Further studies demonstrated that this STAT2-dependent STAT1-independent mechanism requires the type I IFN receptor, and contributes to the autocrine amplification of type I IFN expression. Examination of gene expression in the spleen and bone marrow-derived macrophages following DENV infection revealed STAT2-dependent pathways can induce the transcription of a subset of interferon stimulated genes even in the absence of STAT1. Collectively, these results help elucidate the nature of the poorly understood STAT1-independent host defense mechanism against viruses by identifying a functional type I IFN/STAT2 signaling pathway following DENV infection in vivo.

Important advances in the field of anti-dengue virus research.

There are currently no licensed antivirals available for the treatment of dengue virus (DENV), which causes significant morbidity and mortality throughout tropical areas of the world and is now encroaching on the southern United States. Recent improvements in existing animal models and cell culture systems have been very important in elucidating the mechanisms of DENV pathogenesis in humans, including the identification of potential viral and host proteins that might be targeted for the treatment of DENV infection. The AG129 mouse model is a major advance in the development of antiviral and vaccine candidates for clinical use. It allows for testing of potential therapeutics in a relevant system that exhibits some aspects of disease that are similar to those observed in humans. This review focuses on recent developments in the AG129 mouse model and discusses compounds that have been found to be active in available cell and animal model systems within the past year.

Cardif-mediated signaling controls the initial innate response to dengue virus in vivo.

The role of Cardif-dependent signaling in controlling dengue virus (DENV) infection and regulating type I interferon (IFN) production in vivo was examined in Cardif-deficient mice. DENV RNA levels were significantly elevated in both the serum and lymphoid tissues of Cardif(-/-) mice at early times compared to those in wild-type animals. Type I IFN production was delayed in these locales of Cardif(-/-) mice until 18 h postinfection, indicating that Cardif regulates the initial type I IFN response in lymphoid tissues. In contrast, DENV viral loads in nonlymphoid tissues were similar between Cardif(-/-) and wild-type mice. These results reveal that RNA helicase-mediated sensing acts as a first line of innate defense against DENV infection in vivo and functions in a tissue-dependent manner.

Kaposi's sarcoma-associated herpesvirus virions inhibit interferon responses induced by envelope glycoprotein gpK8.1.

The Kaposi's sarcoma-associated herpesvirus (KSHV) envelope glycoprotein gpK8.1 contributes to cellular attachment through binding cell surface heparan sulfate proteoglycans. By using a soluble recombinant form of gpK8.1, we discovered that a consequence of gpK8.1 interaction with human fibroblasts is the induction of an antiviral response, as characterized by the activation of interferon regulatory factor 3 (IRF-3), production of interferon beta (IFN-beta), and expression of interferon-stimulated antiviral genes. In contrast, neither IFN-beta expression nor a functional antiviral response is observed in cells treated with KSHV virions. The interferon response induced by soluble gpK8.1 can be inhibited by simultaneous treatment with UV-inactivated virions, while the induction of an indicator inflammatory cytokine, interleukin-6, was readily evident in the response to both gpK8.1 and KSHV. In addition, KSHV virions abrogate gpK8.1-mediated activation of IRF-3, an early transcriptional regulator for cellular antiviral responses. Although innate immune responses are initiated during contact between gpK8.1 and cellular receptor(s), these results suggest that the virion contains one or more structural elements that selectively repress an effective antiviral response while allowing cellular responses favorable to the KSHV life cycle.

Human cytomegalovirus elicits a coordinated cellular antiviral response via envelope glycoprotein B.

Previous studies have shown that human cytomegalovirus (CMV) is a potent elicitor of interferon-stimulated gene (ISG) expression. Induction of the interferon pathway does not require replication-competent virus, and envelope glycoprotein B (gB) from CMV is a viral structural component that can directly induce transcription of ISGs. Here we extend these earlier findings by defining the consequences of inducing the interferon pathway. We found that cells respond to CMV or soluble gB by establishing a functional antiviral state within cell types critical in CMV biology, such as fibroblasts and endothelial cells. We have also discovered new insights into the mechanism by which the pathway is initiated. Interferon regulatory factor 3 (IRF3), a key transcriptional regulator of cellular interferon responses, is activated by CMV virions and soluble gB. Thus, IRF3 becomes activated via "outside-in" signal transduction events. This is a novel mechanism of activation of this key transcription factor by viruses. In comparison to soluble gB (gB(1-750)), which comprises the entire ectodomain of gB, a truncation mutant encompassing only the amino-terminal region of gB (gB(1-460)) was markedly less effective at inducing antiviral responses. This indicates that the region of gB from residues 461 to 750 is important for initiation of the antiviral response. In addition, CMV and gB establish an antiviral state in alpha/beta interferon null cells, illustrating that primary induction of ISGs by CMV and gB is sufficient to establish the antiviral response and that interferon secretion is not necessary for the antiviral effect. Taken together, our findings reveal that CMV initiates a coordinated antiviral response through contact between gB and an as-yet-unidentified cell surface receptor(s).