Toward viewing behavior for aerial scene categorization.

Previous work has demonstrated similarities and differences between aerial and terrestrial image viewing. Aerial scene categorization, a pivotal visual processing task for gathering geoinformation, heavily depends on rotation-invariant information. Aerial image-centered research has revealed effects of low-level features on performance of various aerial image interpretation tasks. However, there are fewer studies of viewing behavior for aerial scene categorization and of higher-level factors that might influence that categorization. In this paper, experienced subjects' eye movements were recorded while they were asked to categorize aerial scenes. A typical viewing center bias was observed. Eye movement patterns varied among categories. We explored the relationship of nine image statistics to observers' eye movements. Results showed that if the images were less homogeneous, and/or if they contained fewer or no salient diagnostic objects, viewing behavior became more exploratory. Higher- and object-level image statistics were predictive at both the image and scene category levels. Scanpaths were generally organized and small differences in scanpath randomness could be roughly captured by critical object saliency. Participants tended to fixate on critical objects. Image statistics included in this study showed rotational invariance. The results supported our hypothesis that the availability of diagnostic objects strongly influences eye movements in this task. In addition, this study provides supporting evidence for Loschky et al.'s (Journal of Vision, 15(6), 11, 2015) speculation that aerial scenes are categorized on the basis of image parts and individual objects. The findings were discussed in relation to theories of scene perception and their implications for automation development.

Exploration of drug resistance mechanisms in triple negative breast cancer cells using a microfluidic device and patient tissues.

Chemoresistance is a major cause of treatment failure in many cancers. However, the life cycle of cancer cells as they respond to and survive environmental and therapeutic stress is understudied. In this study, we utilized a microfluidic device to induce the development of doxorubicin-resistant (DOXR) cells from triple negative breast cancer (TNBC) cells within 11 days by generating gradients of DOX and medium. In vivo chemoresistant xenograft models, an unbiased genome-wide transcriptome analysis, and a patient data/tissue analysis all showed that chemoresistance arose from failed epigenetic control of the nuclear protein-1 (NUPR1)/histone deacetylase 11 (HDAC11) axis, and high NUPR1 expression correlated with poor clinical outcomes. These results suggest that the chip can rapidly induce resistant cells that increase tumor heterogeneity and chemoresistance, highlighting the need for further studies on the epigenetic control of the NUPR1/HDAC11 axis in TNBC.

Three-Dimensional Hanging Drop Spheroid Plates for Easy Chimeric Antigen Receptor (CAR) T Cytotoxicity Assay.

Chimeric antigen receptor (CAR) T cell therapy shows a highly effective therapeutic effect on B-cell malignancies. The tumor microenvironment (TME) of solid tumors in vivo poses a great challenge to CAR T cell therapy due to its complexity. Recently, tumor spheroids have attracted much attention because of their ability to recapitulate TME. However, the use of tumor spheroids for the CAR T cytotoxicity assay involves the difficult task of separating unbound T cells and dead tumor cells from the spheroids. Therefore, we developed a three-dimensional hanging spheroid plate (3DHSP) that facilitates spheroid formation and separation of unbound and dead cells from spheroids during cytotoxicity assays. In this work, detailed steps have been described for fabrication and operation of the 3DHSP. This new 3DHSP device is a 96-well plate in which each well consists of a hanging dripper and a spheroid separation plate. A tumor spheroid forms in a droplet hanging in the dripper and is mixed with CAR T cells. The mixture in the droplet is deposited into the spheroid separation plate by pipetting, and unbound and dead CAR T and tumor cells are detached from the spheroid and moved to the waste well in the plate by tilting the 3DHSP at 20°. The size of the spheroid can be used as a readout for CAR T cell cytotoxicity assay, suggesting that the 3DHSP does not require cumbersome fluorescent staining.

Nitisinone attenuates cartilage degeneration and subchondral osteoclastogenesis in osteoarthritis and concomitantly inhibits the cGAS/STING/NF-κB pathway.

Osteoarthritis (OA) is a chronic degenerative joint disease characterized by cartilage degeneration and subchondral bone remodelling. Currently, conservative treatment strategies cannot effectively alleviate the progression of OA. In this study, we used computer network analysis to show that Nitisinone (NTBC) is closely related to extracellular matrix degradation in OA and mainly interferes with the TNF-α signaling pathway. NTBC is an orphan drug used to treat hereditary type I tyrosinemia by altering phenylalanine/tyrosine metabolic flow. In this study, we found that NTBC effectively reduced chondrocyte inflammation and extracellular matrix degradation induced by TNF-α. Mechanistically, NTBC inhibited the cGAS/STING signaling pathway and reduced activation of the STING-dependent NF-κB pathway to alleviate inflammation. In addition, NTBC inhibited osteoclastogenesis and delayed the occurrence of subchondral bone remodelling. In mice with ACLT-induced osteoarthritis, intra-articular injection of NTBC significantly reduced cartilage degradation and subchondral bone remodelling. NTBC showed impressive therapeutic efficacy as a potential pharmaceutical intervention for the treatment of OA.

Exploring Long- and Short-Range Temporal Information for Learned Video Compression.

Learned video compression methods have gained various interests in the video coding community. Most existing algorithms focus on exploring short-range temporal information and developing strong motion compensation. Still, the ignorance of long-range temporal information utilization constrains the potential of compression. In this paper, we are dedicated to exploiting both long- and short-range temporal information to enhance video compression performance. Specifically, for long-range temporal information exploration, we propose a temporal prior that can be continuously supplemented and updated during compression within the group of pictures (GOP). With the updating scheme, the temporal prior can provide richer mutual information between the overall prior and the current frame for the entropy model, thus facilitating Gaussian parameter prediction. As for the short-range temporal information, we propose a progressive guided motion compensation to achieve robust and accurate compensation. In particular, we design a hierarchical structure to build multi-scale compensation, and by employing optical flow guidance, we generate pixel offsets as motion information at each scale. Additionally, the compensation results at each scale will guide the next scale's compensation, forming a flow-to-kernel and scale-by-scale stable guiding strategy. Extensive experimental results demonstrate that our method can obtain advanced rate-distortion performance compared to the state-of-the-art learned video compression approaches and the latest standard reference software in terms of PSNR and MS-SSIM. The codes are publicly available on: https://github.com/Huairui/LSTVC.

Experimental Study and Mechanism Analysis of Paraffin/Sisal Composite Phase Change Energy Storage Fiber Prepared by Vacuum Adsorption Method.

Sisal fiber exhibits a fibrous and porous structure with significant surface roughness, making it highly suitable for storing phase change materials (PCMs). Its intricate morphology further aids in mitigating the risk of PCM leakage. This research successfully employs vacuum adsorption to encapsulate paraffin within sisal fiber, yielding a potentially cost-effective, durable, and environmentally friendly phase change energy storage medium. A systematic investigation was carried out to evaluate the effects of sisal-to-paraffin mass ratio, fiber length, vacuum level, and negative pressure duration on the loading rate of paraffin. The experimental results demonstrate that a paraffin loading rate of 8 wt% can be achieved by subjecting a 3 mm sisal fiber to vacuum adsorption with 16 wt% paraffin for 1 h at -0.1 MPa. Through the utilization of nano-CT imaging enhancement technology, along with petrographic microscopy, this study elucidates the mechanism underlying paraffin storage within sisal fiber during vacuum adsorption. The observations reveal that a substantial portion of paraffin is primarily stored within the pores of the fiber, while a smaller quantity is firmly adsorbed onto its surface, thus yielding a durable phase change energy storage medium. The research findings contribute to both the theoretical foundations and the available practical guidance for the fabrication and implementation of paraffin/sisal fiber composite phase change energy storage mediums.

A 3D hanging spheroid-filter plate for high-throughput drug testing and CAR T cell cytotoxicity assay.

Tumour spheroids are widely used in immune cell cytotoxicity assays and anticancer drug testing, providing a physiologically relevant model replicating the tumour microenvironment. However, co-culture of immune and tumour cells complicates quantification of immune cell killing efficiency. We present a novel 3D hanging spheroid-filter plate that efficiently facilitates spheroid formation and separates unbound/dead cells during cytotoxicity assays. Optical imaging directly measures the cytotoxic effects of anti-cancer drugs on tumour spheroids, eliminating the need for live/dead fluorescent staining. This approach enables cost-effective evaluation of T-cell cytotoxicity with specific chimeric antigen receptors (CARs), enhancing immune cell-based assays and drug testing in three-dimensional tumour models.

Macrophage migration inhibitory factor (MIF) promotes intervertebral disc degeneration through the NF-κB pathway, and the MIF inhibitor CPSI-1306 alleviates intervertebral disc degeneration in a mouse model.

Lumbar intervertebral disc degeneration(IDD) is a prevalent inflammatory disease caused by many proinflammatory factors, such as TNF and IL-1ß. Migration inhibitory factor (MIF) is an upstream inflammatory factor widely expressed in vivo that is associated with a variety of inflammatory diseases or malignant tumors and has potential therapeutic value in many diseases. We explored the role of MIF in intervertebral disc degeneration by regulating the content of exogenous MIF or the expression of MIF in cells. Upon inducing degeneration of nucleus pulposus (NP) cells with IL-1ß, we found that the increase in intracellular and exogenous MIF promoted the catabolism induced by proinflammatory factors in NP cells, while silencing of the MIF gene alleviated the degeneration to some extent. In a mouse model, the intervertebral disc degeneration of MIF-KO mice was significantly less than that of wild-type mice. To explore the treatment of intervertebral disc degeneration, we selected the small-molecular MIF inhibitor CPSI-1306. CPSI-1306 had a therapeutic effect on intervertebral disc degeneration in the mouse model. In summary, we believe that MIF plays an important role in intervertebral disc degeneration and is a potential therapeutic target for the treatment of intervertebral disc degeneration.

Phenotypic assay for cytotoxicity assessment of Balamuthia mandrillaris against human neurospheroids.

Introduction: The phenotypic screening of drugs against Balamuthia mandrillaris, a neuropathogenic amoeba, involves two simultaneous phases: an initial step to test amoebicidal activity followed by an assay for cytotoxicity to host cells. The emergence of three-dimensional (3D) cell cultures has provided a more physiologically relevant model than traditional 2D cell culture for studying the pathogenicity of B. mandrillaris. However, the measurement of ATP, a critical indicator of cell viability, is complicated by the overgrowth of B. mandrillaris in coculture with host cells during drug screening, making it challenging to differentiate between amoebicidal activity and drug toxicity to human cells. Methods: To address this limitation, we introduce a novel assay that utilizes three-dimensional hanging spheroid plates (3DHSPs) to evaluate both activities simultaneously on a single platform. Results and discussion: Our study showed that the incubation of neurospheroids with clinically isolated B. mandrillaris trophozoites resulted in a loss of neurospheroid integrity, while the ATP levels in the neurospheroids decreased over time, indicating decreased host cell viability. Conversely, ATP levels in isolated trophozoites increased, indicating active parasite metabolism. Our findings suggest that the 3DHSP-based assay can serve as an endpoint for the phenotypic screening of drugs against B. mandrillaris, providing a more efficient and accurate approach for evaluating both parasite cytotoxicity and viability.

How do people find pairs?

Humans continuously scan their visual environment for relevant information. Such visual search behavior has typically been studied with tasks in which the search goal is constant and well-defined, requiring relatively little interplay between memory and orienting. Here we studied a situation in which the target is not known in advance, and instead, memory needs to be dynamically updated during the actual search. Observers compared two simultaneously presented arrays of objects for any matching pair of items-a task that requires continuous comparisons between what is seen now and what was seen a few moments ago. To manipulate the balance between memorizing and scanning, we ran two versions of the task. In an eye-tracking version, the objects were continuously available and could be scanned with relative ease. The results suggested that observers preferred scanning over memorizing. In a mouse-tracking version, perceptual availability was limited, and scanning was slowed. Now observers substantially increased their memory use. Thus, the results revealed a flexible and dynamic interplay between memory and perception. The findings aid in further bridging the research fields of attention and memory. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Overexpression of miR-148a-3p inhibits extracellular matrix degradation and alleviates IL-1ß-induced intervertebral disc degeneration.

Objectives: Recently, studies on microRNAs (miRNAs) and their targets and related genes have provided new therapeutic opportunities for controlling intervertebral disc degeneration (IDD). We aimed to investigate the effects of miR-148a-3p overexpression on IDD progression. Materials and Methods: This study used microRNA microarrays to analyze key regulators of IDD. Q-PCR was used to verify the IL-1ß-induced down-regulation of miR-148a-3p expression both in nucleus pulposus (NP) tissues of IDD patients and in degenerated NP cells (NPCs) of rats. Rat NPC micromass cultures and ex vivo intervertebral disc (IVD) culture models were established, and histological staining was performed to verify the effect of miR-148a-3p on the general morphology and proteoglycan and collagen contents of IVDs. In addition, q-PCR and western blotting analyses were performed to examine the expression of ECM molecules and matrix-degrading enzymes. A luciferase reporter assay was used to confirm the target genes of miR-148a-3p. Results: Our data revealed that miR-148a-3p was down-regulated in IDD. Overexpression of miR-148a-3p had no effect on ACAN or COL2A1 gene expression but decreased MMP3, MMP13, and ADAMTS5 gene expression. The matrix deposited by miR-148a-3p-overexpressing rat NPCs contained high levels of proteoglycans and collagen. The ex vivo experiments verified that agomiR-148a-3p alleviated the NPC matrix degradation induced by IL-1ß. A luciferase reporter assay confirmed that miR-148a-3p directly targeted ADAMTS5 and MMP13. Conclusion: We proved that miR-148a-3p can attenuate ECM loss and protect NP function by inhibiting matrix-degrading enzymes.

Cdc42 signaling regulated by dopamine D2 receptor correlatively links specific brain regions of hippocampus to cocaine addiction.

BACKGROUND: Hippocampus plays critical roles in drug addiction. Cocaine-induced modifications in dopamine receptor function and the downstream signaling are important regulation mechanisms in cocaine addiction. Rac regulates actin filament accumulation while Cdc42 stimulates the formation of filopodia and neurite outgrowth. Based on the region specific roles of small GTPases in brain, we focused on the hippocampal subregions to detect the regulation of Cdc42 signaling in long-term morphological and behavioral adaptations to cocaine. METHODS: Genetically modified mouse models of Cdc42, dopamine receptor D1 (D1R) and D2 (D2R) and expressed Cdc42 point mutants that are defective in binding to and activation of its downstream effector molecules PAK and N-WASP were generated, respectively, in CA1 or dentate gyrus (DG) subregion. RESULTS: Cocaine induced upregulation of Cdc42 signaling activity. Cdc42 knockout or mutants blocked cocaine-induced increase in spine plasticity in hippocampal CA1 pyramidal neurons, leading to a decreased conditional place preference (CPP)-associated memories and spatial learning and memory in water maze. Cdc42 knockout or mutants promoted cocaine-induced loss of neurogenesis in DG, leading to a decreased CPP-associated memories and spatial learning and memory in water maze. Furthermore, by using D1R knockout, D2R knockout, and D2R/Cdc42 double knockout mice, we found that D2R, but not D1R, regulated Cdc42 signaling in cocaine-induced neural plasticity and behavioral changes. CONCLUSIONS: Cdc42 acts downstream of D2R in the hippocampus and plays an important role in cocaine-induced neural plasticity through N-WASP and PAK-LIMK-Cofilin, and Cdc42 signaling pathway correlatively links specific brain regions (CA1, dentate gyrus) to cocaine-induced CPP behavior.

Recent Progress in Biosensors for Detection of Tumor Biomarkers.

Cancer is a leading cause of death worldwide, with an increasing mortality rate over the past years. The early detection of cancer contributes to early diagnosis and subsequent treatment. How to detect early cancer has become one of the hot research directions of cancer. Tumor biomarkers, biochemical parameters for reflecting cancer occurrence and progression have caused much attention in cancer early detection. Due to high sensitivity, convenience and low cost, biosensors have been largely developed to detect tumor biomarkers. This review describes the application of various biosensors in detecting tumor markers. Firstly, several typical tumor makers, such as neuron-specific enolase (NSE), carcinoembryonic antigen (CEA), prostate-specific antigen (PSA), squamous cell carcinoma antigen (SCCA), carbohydrate, antigen19-9 (CA19-9) and tumor suppressor p53 (TP53), which may be helpful for early cancer detection in the clinic, are briefly described. Then, various biosensors, mainly focusing on electrochemical biosensors, optical biosensors, photoelectrochemical biosensors, piezoelectric biosensors and aptamer sensors, are discussed. Specifically, the operation principles of biosensors, nanomaterials used in biosensors and the application of biosensors in tumor marker detection have been comprehensively reviewed and provided. Lastly, the challenges and prospects for developing effective biosensors for early cancer diagnosis are discussed.

Taohong Siwu Decoction Promotes Osteo-Angiogenesis in Fractures by Regulating the HIF-1α Signaling Pathway.

Background: Vascular damage is a major consequence of bone fracture. Taohong Siwu decoction (TSD) can raise the expression of vascular endothelial growth factor (VEGF) in fracture healing. However, its molecular mechanism in promoting angiogenesis is still unknown. The aim of this study was to investigate the potential mechanisms of TSD in the regulation of osteo-angiogenesis in fracture healing. Methods: A rat tibial fracture model was established. After low- (4.5 g·kg-1), medium- (9 g·kg-1), and high-dose TSD (18 g·kg-1) and panax notoginsenoside (25 mg kg-1) treatment, hematoxylin-eosin staining was employed to visualize pathological changes in bone tissues. The levels of cytokines (interleukin (IL)-2, tumor necrosis factor-α (TNF-α), IL-6, and IL-1ß), thromboxane B2 (TXB2), and 6 ketone prostaglandin F1α (6-Keto-PGF1α) were quantified by enzyme-linked immunosorbent assay (ELISA). Immunofluorescence was used to identify the rat aortic endothelial cells (RAECs). Control serum, 10% TSD-containing serum, and 10% TSD-containing serum combined with hypoxia-inducible factor-1α (HIF-1α) inhibitor were used to treat the RAECs and rat osteoblasts. Transwell migration assay was utilized to examine the migration of the RAECs. The Matrigel tubulogenesis assay was used for the assessment of angiogenesis. The expression of angiogenesis- (von Hippel-Lindau tumor suppressor (VHL), HIF-1α, VEGF, angiopoietin-2 (Ang-2), and pVHL) and osteogenesis-related (alkaline phosphatase (ALP), runt-related transcription factor 2 (Runx2), and osteopontin-1 (OPN-1)) protein and gene was detected by western blot and quantitative real-time PCR (qRT-PCR). Results: Compared with the model group, TSD increased the trabecular bone areas, numbers, and thicknesses in fractured rats. In the plasma, the levels of cytokines and TXB2 in the middle- and high-dose TSD group were significantly lower than those in the model group (P < 0.01). The 6-keto-PGF1α content was increased by middle- and high-dose TSD intervention (P < 0.01). Compared to the control serum group, the angiogenesis and migration of the RAECs were enhanced in the TSD group (P < 0.001). The expression of HIF-1α, VEGF, and Ang-2 in the TSD group upregulated significantly (P < 0.001). VHL and pVHL were inhibited under TSD-containing serum treatment (P < 0.001). ALP, Runx2, and OPN-1 were increased obviously in the TSD group (P < 0.001). Nevertheless, the HIF-1α inhibitor reversed these changes (P < 0.001). Conclusion: TSD promotes angiogenesis and osteogenesis by regulating the HIF-1α signaling pathway. Meanwhile, it can effectively reduce the risk of inflammation and improve blood circulation.

High-Throughput Screening of Anti-cancer Drugs Using a Microfluidic Spheroid Culture Device with a Concentration Gradient Generator.

Tumor spheroid models are widely used for drug screening as in vitro models of the tumor microenvironment. There are various ways in which tumor spheroid models can be prepared, including the self-assembly of cells using low-adherent plates, micro-patterned plates, or hanging-drop plates. Recently, drug high-throughput screening (HTS) approaches have incorporated the use of these culture systems. These HTS culture systems, however, require complicated equipment, such as robot arms, detectors, and software for handling solutions and data processing. Here, we describe protocols that allow tumor spheroids to be tested with different concentrations of a drug in a parallel fashion using a microfluidic device that generates a gradient of anti-cancer drugs. This microfluidic spheroid culture device with a concentration gradient generator (µFSCD-CGG) enables the formation of 50 tumor spheroids and the testing of drugs at five different concentrations. First, we provide a protocol for the fabrication of the µFSCD-CGG, which has both a culture array in which tumor cells are injected and aggregate to form spheroids and a concentration gradient generator for drug testing. Second, we provide a protocol for tumor spheroid formation and HTS of anti-cancer drugs using the device. Finally, we provide a protocol for assessing the response of tumor spheroids at different drug concentrations. To address the needs of the pharmaceutical industry, this protocol can be used for various cell types, including stem cells, and the number of tumor spheroids and drug concentration ranges that can be tested in the µFSCD-CGG can be increased. © 2022 Wiley Periodicals LLC. Basic Protocol 1: Fabrication of a microfluidic spheroid culture device with a concentration gradient generator (µFSCD-CGG) Basic Protocol 2: Seeding cells and formation of spheroids in the µFSCD-CGG Basic Protocol 3: Drug treatment and assessment of cell viability in the µFSCD-CGG.

TRAF4 Maintains Deubiquitination of Caveolin-1 to Drive Glioblastoma Stemness and Temozolomide Resistance.

Glioblastoma (GBM) is the most common type of primary adult brain tumor. Glioma stem cell (GSC) residence and temozolomide (TMZ) resistance in GBM both contribute to poor patient outcome. TRAF4 is a scaffold protein with E3 ubiquitin ligase activity that has recently been discovered to promote invasion and metastasis in several malignancies, but the effects and functions of TRAF4 in GBM remain to be determined. Here, we report that TRAF4 is preferentially overexpressed in GSCs and is required for stem-like properties as well as TMZ sensitivity in GBM cells. TRAF4 specifically interacted with the N-terminal tail of Caveolin-1 (CAV1), an important contributor to the tumorigenicity of GBM cells. TRAF4 regulated CAV1 stability by preventing ZNRF1-mediated ubiquitination and facilitating USP7-mediated deubiquitination independently of its E3 ubiquitin ligase catalytic activity. TRAF4-mediated stabilization of CAV1 activated protumorigenic AKT/ERK1/2 signaling, and disruption of this axis resulted in defects in stemness maintenance. In addition, expression of TRAF4 and CAV1 was positively correlated and predicted poor prognosis in human GBM samples. Screening of common nervous system drugs identified risperidone interaction with TRAF4, and risperidone treatment resulted in the dissociation of TRAF4 and CAV1. Importantly, pharmacologic inhibition of TRAF4 with risperidone potently inhibited self-renewal, abrogated tumorigenicity, and reversed TMZ resistance in GBM. Overall, TRAF4-mediated stabilization of CAV1 promotes stemness and TMZ resistance in GBM, providing a therapeutic strategy that could improve patient outcomes. SIGNIFICANCE: The identification of a TRAF4/Caveolin-1 axis that plays a crucial role in malignant progression of glioblastoma provides new insights into the function of TRAF4 in ubiquitin signaling and suggests TRAF4 as a potential therapeutic target.

Carboxylated fluorescent microsphere based immunochromatographic test strip enabled sensitive and quantitative on-site detection for florfenicol in eggs.

Florfenicol (FF), used popularly in prevention and treatment of virus infections in livestock and poultry, has widely been found in eggs and harmful to human health. In this work, a sensitive and quantitative on-site detecting solution, monoclonal antibody-based carboxylated fluorescent microsphere immunochromatographic test strip assay (FM-ICTS), is design and applied for FF detection. The proposed method can sensitively detect FF in low detection limit of 0.030 ng/g and quantitatively measure its concentration from 0.1 ng/mL to 8.1 ng/mL (R2 = 0.9991) with high repeatability (CV<8.0 %). In addition, the established FM-ICTS method exhibited high measurement accuracy in FF samples as compared with HPLC-MS analysis and demonstrated satisfied recoveries (99.1-101.3 %). More importantly, the quantitative FF test strip demonstrate ultra-high stability, which presents approximately equivalent detection ability to the fresh one after stored at 4 °C for more than one year or stored at 37 °C for 60 days. Therefore, the proposed method is a promising solution for rapidly and sensitively quantitative determination of FF in eggs.

Predicate Correlation Learning for Scene Graph Generation.

For a typical Scene Graph Generation (SGG) method in image understanding, there usually exists a large gap in the performance of the predicates' head classes and tail classes. This phenomenon is mainly caused by the semantic overlap between different predicates as well as the long-tailed data distribution. In this paper, a Predicate Correlation Learning (PCL) method for SGG is proposed to address the above problems by taking the correlation between predicates into consideration. To measure the semantic overlap between highly correlated predicate classes, a Predicate Correlation Matrix (PCM) is defined to quantify the relationship between predicate pairs, which is dynamically updated to remove the matrix's long-tailed bias. In addition, PCM is integrated into a predicate correlation loss function ( LPC ) to reduce discouraging gradients of unannotated classes. The proposed method is evaluated on several benchmarks, where the performance of the tail classes is significantly improved when built on existing methods.

3D hanging spheroid plate for high-throughput CAR T cell cytotoxicity assay.

BACKGROUND: Most high-throughput screening (HTS) systems studying the cytotoxic effect of chimeric antigen receptor (CAR) T cells on tumor cells rely on two-dimensional cell culture that does not recapitulate the tumor microenvironment (TME). Tumor spheroids, however, can recapitulate the TME and have been used for cytotoxicity assays of CAR T cells. But a major obstacle to the use of tumor spheroids for cytotoxicity assays is the difficulty in separating unbound CAR T and dead tumor cells from spheroids. Here, we present a three-dimensional hanging spheroid plate (3DHSP), which facilitates the formation of spheroids and the separation of unbound and dead cells from spheroids during cytotoxicity assays. RESULTS: The 3DHSP is a 24-well plate, with each well composed of a hanging dripper, spheroid wells, and waste wells. In the dripper, a tumor spheroid was formed and mixed with CAR T cells. In the 3DHSP, droplets containing the spheroids were deposited into the spheroid separation well, where unbound and dead T and tumor cells were separated from the spheroid through a gap into the waste well by tilting the 3DHSP by more than 20°. Human epidermal growth factor receptor 2 (HER2)-positive tumor cells (BT474 and SKOV3) formed spheroids of approximately 300-350 µm in diameter after 2 days in the 3DHSP. The cytotoxic effects of T cells engineered to express CAR recognizing HER2 (HER2-CAR T cells) on these spheroids were directly measured by optical imaging, without the use of live/dead fluorescent staining of the cells. Our results suggest that the 3DHSP could be incorporated into a HTS system to screen for CARs that enable T cells to kill spheroids formed from a specific tumor type with high efficacy or for spheroids consisting of tumor types that can be killed efficiently by T cells bearing a specific CAR. CONCLUSIONS: The results suggest that the 3DHSP could be incorporated into a HTS system for the cytotoxic effects of CAR T cells on tumor spheroids.

Learning Discrete Representations From Reference Images for Large Scale Factor Image Super-Resolution.

Image super-resolution (SR) task aims to recover high-resolution (HR) images from degraded low-resolution (LR) images, which has achieved great progress due to the recent advances of deep neural networks. Due to severe information loss of the LR images, it is more challenging to reconstruct high quality HR images at large scale factors, i. e., higher than 4× . Traditional reference image based SR methods usually perform patch matching to locate detailed texture from HR reference images which could provide fine details from similar image contents. But it suffers from difficulties in achieving good matching in the largely downscaled image space or feature space due to the ill-posed nature between LR and HR mapping. In this paper, we tackle this problem by exploiting fine details contained in reference HR images. Inspired by vector quantization (VQ), we propose a simple yet effective auto-encoder convolutional neural network (CNN) module to learn discrete representations of images. Furthermore, we propose to progressively learn pairs of cross-scale discrete feature representations using paired LR and HR reference images. The coarser scale of the discrete representation is responsible for encoding the global image structure while the paired finer scale of the discrete representation takes charge of capturing missing details in the finer image scale. During inference, continuous features of the test LR image are used as queries to retrieve finer scale discrete representations (value) by searching the nearest coarser scale discrete representations (key). Then, the queries and retrieved values are combined to progressively recover the HR image. Experimental results indicate that when compared with the state-of-the-art image SR models, the proposed method can achieve advanced performance in terms of both objective quality and subjective quality. The code will be available on URL: https://github.com/sunwj/refsr.