An optofluidic system for the concentration gradient screening of oocyte protection drugs.

Oocytes protective drug screening is essential for the treatment of reproductive diseases. However, few studies construct the oocyte in vitro drug screening microfluidic systems because of their enormous size, scarcity, and sensitivity to the culture environment. Here, we present an optofluidic system for oocyte drug screening and state analysis. The system consists of two parts: an open-top drug screening microfluidic chip and an optical Fourier filter analysis part. The open-top microfluidic chip anchors single oocyte with hydrogel and allows nutrient and gas environment updating which is essential for oocyte culturing. The optical filter analysis part is used to accurately analyse the status of oocytes. Based on this system, we found that fluorene-9-bisphenol (BHPF) damaged the oocyte spindle in a dose-dependent manner, a high dose of melatonin (10-3 M) effectively reduces the percentage of abnormally arranged chromosomes of oocytes exposed to 40 µM BHPF. This optofluidic system shows great promise for the culture of oocytes and demonstrates the robust ability for convenient multi-concentration oocytes drug screening. This technology may benefit further biomedicine and reproductive toxicology applications in the lab on a chip community.

Fabrication and Performance of Bubble-Containing Multicompartmental Particles: Novel Self-Orienting Carriers.

In this work, a class of bubble-containing multicompartmental particles with self-orienting capability is developed, where a single bubble is enclosed at the top of the super-segmented architecture. Such bubbles, driven by potential energy minimization, cause the particles to have a bubble-upward preferred orientation in liquid, enabling efficient decoding of their high-density signals in an interference-resistant manner. The particle preparation involves bubble encapsulation via the impact of a multicompartmental droplet on the liquid surface and overall stabilization via rational crosslinking. The conditions for obtaining these particles are systematically investigated. Methodological compatibility with materials is demonstrated by different hydrogel particles. Finally, by encapsulating cargoes of interest, these particles have found broad applications in actuators, multiplexed detection, barcodes, and multicellular systems.

The preventive effect of dexmedetomidine on anesthesia complications in strabismus surgery: a systematic review and meta-analysis.

OBJECTIVE: Dexmedetomidine is a medication that has analgesic, sedative, and anti-anxiety properties. In the clinical, it is often used to prevent common complications associated with strabismus surgery, including postoperative delirium, postoperative nausea and vomiting, postoperative pain, and oculocardiac reflex. However, its effectiveness and side effects of the present studies are different. The sample sizes of the present studies on the prevention of complications of dexmedetomidine are small. Therefore, this study evaluates the efficacy of dexmedetomidine in preventing anesthesia-related complications in strabismus surgery through a systematic review and meta-analysis. METHODS: Literature was retrieved from 10 commonly used databases and randomized controlled trials published up to May 2022 were sought. The included studies compared the intervention effects of dexmedetomidine versus placebo on anesthesia-related complications in surgery. The occurrence rates of postoperative delirium, postoperative nausea and vomiting, postoperative pain, and oculocardiac reflex in patients undergoing strabismus surgery were evaluated. Statistical analyses and forest plots were generated using Review Manager and STATA software. Binary outcomes were measured using relative risk (RR) with a 95% confidence interval for each outcome. The Cochrane risk of bias tool was used to assess the bias and risk in the studies that met the inclusion criteria. RESULTS: A total of 13 articles were ultimately included in the analysis, comprising 1,018 patients who underwent strabismus surgery. The dexmedetomidine group, compared to the placebo group, demonstrated significant reductions in the incidence of postoperative delirium (RR = 0.73, P = 0.001), severe postoperative delirium (RR = 0.45, P = 0.005), postoperative nausea and vomiting (RR = 0.48, P < 0.0001), and the need for supplemental analgesia postoperatively (RR = 0.60, P = 0.004). Additionally, subgroup analysis revealed that intravenous administration of dexmedetomidine significantly reduced the incidence of oculocardiac reflex (RR = 0.50, P = 0.001). In contrast, intranasal administration of dexmedetomidine did not have a significant effect on the incidence of oculocardiac reflex (RR = 1.22, P = 0.15). There was a significant difference between the subgroups (P = 0.0005, I2 = 91.7%). CONCLUSION: Among patients undergoing strabismus surgery, the use of dexmedetomidine can alleviate postoperative delirium and reduce the incidence of postoperative nausea and vomiting, as well as postoperative pain. Moreover, intravenous administration of dexmedetomidine can lower the occurrence rate of the oculocardiac reflex.

Stability of the Arginine-Ornithine-Citrulline Cycle Maintained by Tumor-Stroma Interactions in Cell Coculture Hydrogel Microspheres.

Tumor-stroma interaction is the core process of tumor metastasis. Therefore, building a model of tumor-stromal cell communication is crucial for understanding the tumor metastasis process and curing cancer completely. In this research, a controllable three-dimensional (3D) tumor-stroma coculture microsphere model based on droplet microfluidic technology was developed to culture human lung cancer cells (A549 cell) and fibroblast cells (NIH-3T3 cell) using core-shell hydrogel microspheres to partition different kinds of cells. In our coculture model, tumor cells show a trend of epithelial-mesenchymal transition (EMT): a decrease in the number of surface E-cadherin and an increase in the number of N-cadherin. At the same time, fibroblasts are activated into cancer-associated fibroblasts (CAFs) as the level of interleukin-6 (IL-6) released is increased. In addition, an interesting phenomenon was discovered; in the absence of fibroblasts, the metabolism of the tumor cell culture alone leads to arginine depletion and citrulline accumulation, whereas a coculture can maintain the arginine-ornithine-citrulline cycle to reach equilibrium after 72 h, and the balance increases the stress resistance of tumor cells. This discovery may provide a new direction for understanding tumor resistance.

Multicompartmental Hydrogel Microspheres as a Tool for Multicomponent Analysis.

Developing advanced tools for multicomponent analysis is an open challenge in engineering and life science. Herein, multicompartmental hydrogel microspheres with multi-material compatibility and structural scalability are developed as a tool for multicomponent analysis at a single-particle level. Microfluidic technology endows particles with adjustable sizes and super-segmented layouts that can be used to load various analyte probes. In order to perform multicomponent analysis, these microspheres are structurally divided into identifier regions for indicating reading direction and analyte regions for detecting target molecules. The multiplex detection ability of these particles is demonstrated in microRNA bioassays with high specificity and sensitivity. The multi-target analysis is performed on a single-particle level, and the bioassay is free of conventional labeling interference. We expect these particles to reach their potential in clinical diagnostics.

In situ polymerized ionic liquids in polyester fiber composite membranes for detection of trace oil.

In situ trace detection on ultra-clean surfaces is an important technology. The polyester fiber (PF) was introduced to serve as the template, to which the ionic liquids were bonded by hydrogen bonding. Polymerized ionic liquids (PIL) in PF were formed by in situ polymerization with the azodiisobutyronitrile (AIBN) and IL. The trace oil on metal surfaces was enriched by the composite membrane based on similar compatibility principle. The absolute recovery of the trace oil ranged from 91%-99% using this composite membrane. In the extraction samples, desirable linear correlations were obtained for trace oil in the range of 1.25-20 mg/mL. It has been proven that a 1 cm2 PIL-PF composite membrane can effectively extract as little as 1 mg of lubricating oil on an ultra-clean metal surface of 0.1 m2 with the LOD of 0.9 mg/mL, making it a promising material for in situ detection of trace oil on metal surfaces.

Multidimensional controllable fabrication of tumor spheroids based on a microfluidic device.

Multicellular tumor spheroids (MCTSs) are in vitro solid tumor models with physiological relevance. To achieve robust process control, a MCTS fabrication method that combines cell membrane engineering and droplet microfluidic techniques is designed. The fluidic control and the chemical interactions between biotin and streptavidin enable artificial cell aggregation to be accomplished in seconds. Then, spheroids with a uniform size are fabricated within alginate microcapsules. Microfluidic mixing-based cell aggregation regulates the cell aggregate size and the spheroid composition, and the microcapsules regulate the size of spheroids from 120 to 180 µm. The method shows applicability for various cancer cell lines, including HCT116, HepG2, and A549. In addition, composite colon cancer spheroids consisting of HCT116 and NIH3T3 with predetermined cell ratios and uniform distributions are produced. The generated MCTSs are assessed using the ELISA and UPLC-MS/MS techniques. The release of vascular endothelial growth factor (VEGF) and the 5-fluorouracil (5-FU) resistance differ in the monotypic and cocultured colon cancer models. Our method provides a robust way to produce consistent and customized MCTSs in cancer research and drug screening.

Capture the Moment: High-Speed Imaging With Spiking Cameras Through Short-Term Plasticity.

High-speed imaging can help us understand some phenomena that are too fast to be captured by our eyes. Although ultra-fast frame-based cameras (e.g., Phantom) can record millions of fps at reduced resolution, they are too expensive to be widely used. Recently, a retina-inspired vision sensor, spiking camera, has been developed to record external information at 40, 000 Hz. The spiking camera uses the asynchronous binary spike streams to represent visual information. Despite this, how to reconstruct dynamic scenes from asynchronous spikes remains challenging. In this paper, we introduce novel high-speed image reconstruction models based on the short-term plasticity (STP) mechanism of the brain, termed TFSTP and TFMDSTP. We first derive the relationship between states of STP and spike patterns. Then, in TFSTP, by setting up the STP model at each pixel, the scene radiance can be inferred by the states of the models. In TFMDSTP, we use the STP to distinguish the moving and stationary regions, and then use two sets of STP models to reconstruct them respectively. In addition, we present a strategy for correcting error spikes. Experimental results show that the STP-based reconstruction methods can effectively reduce noise with less computing time, and achieve the best performances on both real-world and simulated datasets.

Multivariate Porous Aromatic Frameworks with High Porosity and Hierarchical Structures for Enzyme Immobilization.

As materials with permanently porous structures and readily modifying availability, porous aromatic frameworks (PAFs) are considered as promising porous materials with versatile functionality. Currently the designable synthesis of PAFs with the desired surface area and pore size is still a challenge, and instead kinetically irreversible coupling reactions for PAFs synthesis has resulted in the unpredictable connection of building units. Herein, a series of PAFs with highly porous and hierarchical structures were successfully synthesized through a multivariate inspired strategy, where multiple building units with various topologies and sizes were selected for PAFs synthesis. All the PAFs synthesized through this strategy possessed hierarchical structures and high specific surface areas at the same time. Encouraged by their high surface area and hierarchical structures, we loaded lipase onto one of the multivariate PAFs. The enzyme loading content of the obtained lipase@PAF-147 was as high as 1456 mg g-1, which surpassed any other currently reported enzyme loading materials. The lipase@PAF-147 also exhibited favorable catalytic activity and stability to a model reaction of p-nitrophenyl caprylate (p-NPC) hydrolysis. This multivariate strategy inspired synthetic method broadens the selection of building units for PAFs design and opens a new avenue for the design of functional porous materials.

Microfluidic Engineering of Crater-Terrain Hydrogel Microparticles: Toward Novel Cell Carriers.

Fabrication and application of novel anisotropic microparticles are of wide interest. Herein, a new method for producing novel crater-terrain hydrogel microparticles is presented using a concept of droplet-aerosol impact and regional polymerization. The surface pattern of microparticles is similar to the widespread "crater" texture on the lunar surface and can be regulated by the impact morphology of aerosols on the droplet surface. Methodological applicability was demonstrated by producing ionic-cross-linked (alginate) and photo-cross-linked (poly(ethylene glycol) diacrylate, PEGDA) microparticles. Additionally, the crater-terrain microparticles (CTMs) can induce nonspecific protein absorption on their surface to acquire cell affinity, and they were exploited as cell carriers to load living cells. Cells could adhere and proliferate, and a special cellular adhesion fingerprint was observed on the novel cell carrier. Therefore, the scalable manufacturing method and biological potential make the engineered microparticles promising to open a new avenue for exploring cell-biomaterial crosstalk.

Deletion of Insulin-like growth factor II mRNA-binding protein 3 participates in the pathogenesis of recurrent spontaneous abortion by inhibiting IL-10 secretion and inducing M1 polarization.

Insulin-like growth factor II mRNA-binding protein 3 (IGF2BP3) has been proved to affect trophoblast function and embryonic development, but its role and potential mechanism in recurrent spontaneous abortion (RSA) are not clear. RSA is a complex reproductive disease, causing physical and mental damage to patients. In recent years, many studies have found that immune microenvironment is vital to maintain successful pregnancy in the maternal fetal interface. Therefore, this study aims to explore the role of IGF2BP3 in affecting macrophage polarization and its possible mechanism. In this article, we found that IGF2BP3 expression was decreased in placental villous samples of human and RSA mouse model, and knockdown of IGF2BP3 in HTR8/SVneo cells promotes M1 Mφ polarization. Combining with RNA sequencing analysis, we found that IGF2BP3 may regulate the Mφ polarization by affecting the expression of trophoblast cytokines, especially IL-10 secretion. Further mechanistic studies showed that knockdown of IGF2BP3 decreased expression of IL-10 by activating NF-κB pathway. Moreover, we found that M2 Mφ promote trophoblast invasion not IGF2BP3 dependent. Our study reveals the interaction between trophoblast cells and macrophages at the maternal-fetal interface of RSA patients, and will provide theoretical guidance for its diagnosis and treatment of RSA patients.

YY1/ITGA3 pathway may affect trophoblastic cells migration and invasion ability.

Recurrent spontaneous abortion (RSA) is a disturbing pregnancy disorder experienced by ~2.5% of women attempting to conceive. The pathogenesis of RSA is still unclear. Previous findings revealed that transcription factor YIN-YANG 1(YY1) was related to the pathogenesis of RSA by influence trophoblastic cell invasion ability. Present study aimed to investigate more specific molecular mechanism of YY1 playing in trophoblastic cells. In our research, RNA-seq and Chip-seq were used to find significant changed genes between si-YY1(Knock down of YY1) HTR-8/SVneo cells(n = 3) and HTR-8/SVneo cells(n = 3). Gene Ontology and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis results suggested that Integrins related pathway maybe necessary to biological functions of trophoblastic cells. Chip-seq dataset analysis results predict YY1 can regulate ITGA3/7 expression by binding to the promoter region of ITGA3/7. Furthermore, results from chip experiment, RT-PCR, Dual-luciferase reporter gene assay showed that YY1 was able to bind to the promoter region of ITGA3 and regulate ITGA3 mRNA and protein expression. However, ITGA7 could not be significant influenced by YY1. Besides, gene silencing experiment, Western blot and Immunofluorescence assay confirmed that both YY1 and ITGA3 can accelerate phosphorylation focal adhesion kinase and affect cytoskeleton formation in HTR-8/SVneo cells. In conclusion, YY1/ITGA3 play a critical role in trophoblast invasion ability by regulating cytoskeleton formation.

The role of Sirtuin 1 in the pathophysiology of polycystic ovary syndrome.

Polycystic ovarian syndrome (PCOS) is the most common multifactor heterogeneous endocrine and metabolic disease in women of childbearing age. PCOS is a group of clinical syndromes characterized by reproductive disorders, metabolic disorders, and mental health problems that seriously impact the physical and mental health of patients. At present, new studies suggest that human evolution leads to the body changes and the surrounding environment mismatch adaptation, but the understanding of the disease is still insufficient, the pathogenesis is still unclear. Sirtuin 1 (SIRT1), a member of the Sirtuin family, is expressed in various cells and plays a crucial role in cell energy conversion and physiological metabolism. Pathophysiological processes such as cell proliferation and apoptosis, autophagy, metabolism, inflammation, antioxidant stress and insulin resistance play a crucial role. Moreover, SIRT1 participates in the pathophysiological processes of oxidative stress, autophagy, ovulation disturbance and insulin resistance, which may be a vital link in the occurrence of PCOS. Hence, the study of the role of SIRT1 in the pathogenesis of PCOS and related complications will contribute to a more thorough understanding of the pathogenesis of PCOS and supply a basis for the treatment of patients.

Integrated Bioinformatic Analysis of DNA Methylation and Immune Infiltration in Endometrial Cancer.

Background: Endometrial cancer greatly threatens the health of female. Emerging evidences have demonstrated that DNA methylation and immune infiltration are involved in the occurrence and development of endometrial cancer. However, the mechanism and prognostic biomarkers of endometrial cancer are still unclear. In this study, we assess DNA methylation and immune infiltration via bioinformatic analysis. Methods: The latest RNA-Seq, DNA methylation data, and clinical data related to endometrial cancer were downloaded from the UCSC Xena dataset. The methylation-driven genes were selected, and then the risk score was obtained using "MethylMix" and "corrplot" R packages. The connection between methylated genes and the expression of screened driven genes were explored using "survminer" and "beeswarm" packages, respectively. Finally, the role of VTCN1in immune infiltration was analyzed using "CIBERSORT" package. Results: In this study, 179 upregulated genes, and 311 downregulated genes were identified and found to be related to extracellular matrix organization, cell-cell junctions, and cell adhesion molecular binding. The methylation-driven gene VTCN1 was selected, and patients classified to the hypomethylation and high expression group displayed poor prognosis. The VTCN1 gene exhibited highest correlation coefficient between methylation and expression. More importantly, the hypomethylation of promoter of VTCN1 led to its high expression, thereby induce tumor development by inhibiting CD8+ T cell infiltration. Conclusions: Overall, our study was the first to reveal the mechanism of endometrial cancer by assessing DNA methylation and immune infiltration via integrated bioinformatic analysis. In addition, we found a pivotal prognostic biomarker for the disease. Our study provides potential targets for the diagnosis and prognosis of endometrial cancer in the future.

Irisin reduces the abnormal reproductive and metabolic phenotypes of PCOS by regulating the activity of brown adipose tissue in mice†.

Polycystic ovary syndrome (PCOS) is a common endocrine and metabolic disease in women, with clinical manifestations of anovulation and hyperandrogenaemia. The treatment of PCOS mainly focuses on improving clinical symptoms, such as insulin sensitivity or menstrual disorder, through drug treatment. However, due to the pathogenesis diversity of PCOS, there is still a lack of effective treatment in clinics. Metabolic disorder is the key factor in the occurrence of PCOS. Brown adipose tissue (BAT) is a special adipose tissue in the human body that can participate in metabolic balance by improving heat production. BAT has been demonstrated to be an important substance involved in the metabolic disorder of PCOS. Although increasing evidence indicates that BAT transplantation can improve the symptoms of PCOS, it is difficult to achieve BAT transplantation at present due to technical limitations. Stimulation of BAT activation by exogenous substances may be an effective alternative therapy for PCOS. In this study, we investigated the effects of Irisin on dehydroepiandrosterone (DHEA)-induced PCOS in mice and evaluated the effect of Irisin on serum hormone levels and changes in body temperature, body weight, and ovarian morphology. In our study, we found that Irisin can enhance the thermogenesis and insulin sensitivity of PCOS mice by activating the function of BAT. In addition, Irisin treatment can correct the menstrual cycle of PCOS mice, improve the serum steroid hormone disorder status, and reduce the formation of ovarian cystic follicles. In conclusion, our results showed that Irisin treatment significantly improved the metabolic disorder of PCOS and may provide a new and alternative therapy for the treatment of this pathology.

Global methane and nitrous oxide emissions from inland waters and estuaries.

Inland waters (rivers, reservoirs, lakes, ponds, streams) and estuaries are significant emitters of methane (CH4 ) and nitrous oxide (N2 O) to the atmosphere, while global estimates of these emissions have been hampered due to the lack of a worldwide comprehensive data set of CH4 and N2 O flux components. Here, we synthesize 2997 in-situ flux or concentration measurements of CH4 and N2 O from 277 peer-reviewed publications to estimate global CH4 and N2 O emissions from inland waters and estuaries. Inland waters including rivers, reservoirs, lakes, and streams together release 95.18 Tg CH4  year-1 (ebullition plus diffusion) and 1.48 Tg N2 O year-1 (diffusion) to the atmosphere, yielding an overall CO2 -equivalent emission total of 3.06 Pg CO2  year-1 . The estimate of CH4 and N2 O emissions represents roughly 60% of CO2 emissions (5.13 Pg CO2  year-1 ) from these four inland aquatic systems, among which lakes act as the largest emitter for both CH4 and N2 O. Ebullition showed as a dominant flux component of CH4 , contributing up to 62%-84% of total CH4 fluxes across all inland waters. Chamber-derived CH4 emission rates are significantly greater than those determined by diffusion model-based methods for commonly capturing of both diffusive and ebullitive fluxes. Water dissolved oxygen (DO) showed as a dominant factor among all variables to influence both CH4 (diffusive and ebullitive) and N2 O fluxes from inland waters. Our study reveals a major oversight in regional and global CH4 budgets from inland waters, caused by neglecting the dominant role of ebullition pathways in those emissions. The estimated indirect N2 O EF5 values suggest that a downward refinement is required in current IPCC default EF5 values for inland waters and estuaries. Our findings further indicate that a comprehensive understanding of the magnitude and patterns of CH4 and N2 O emissions from inland waters and estuaries is essential in defining the way of how these aquatic systems will shape our climate.

The Mechanism of Insulin-Like Growth Factor II mRNA-Binging Protein 3 Induce Decidualization and Maternal-Fetal Interface Cross Talk by TGF-ß1 in Recurrent Spontaneous Abortion.

Recurrent spontaneous abortion (RSA) is defined as the loss of two or more consecutive intrauterine pregnancies that are clinically established early in pregnancy. To date, the etiology and underlying mechanisms of RSA remain unclear. It is widely thought that the impairment of decidualization is inclined to induce subsequent pregnancy failure and leads to the dysregulation of extra-villous trophoblast invasion and proliferation through maternal-fetal cross talk. However, the mechanism of decidualization in RSA has yet to be understood. In our study, we demonstrate that decidual samples from RSA patients have significantly higher insulin-like growth factor 2 mRNA-binding protein 3 (IGF2BP3) and lower TGF-ß1 levels compared to healthy controls. In addition, the overexpression of IGF2BP3 in human endometrial stromal cells (hESCs) can lead to the impairment of decidualization in vitro-induced model and the abnormal cell cycle regulation. Furthermore, TGF-ß1 and MMP9 levels were greatly increased after decidualization, whereas IGF2BP3 overexpression inhibited endometrial mesenchymal decidualization by downregulating TGF-ß1, impeding maternal-fetal interface cytokine cross talk, and limiting the ability of trophoblast invasion. In conclusion, our investigation first demonstrates that abnormal elevation of IGF2BP3 in the pregnant endometrium leads to the impairment of decidualization and abnormal trophoblast invasion, thereby predisposing individuals to RSA.

Microfluidic-based in vitro thrombosis model for studying microplastics toxicity.

The potential impact of microplastics (MPs) on health has caused great concern, and a toxicology platform that realistically reproduces the system behaviour is urgently needed to further explore and validate MP-related health issues. Herein, we introduce an optically assisted thrombus platform to reveal the interaction of MPs with the vascular system. The risk of accumulation has also been evaluated using a mouse model, and the effect of MPs on the properties of the thrombus are validated via in vitro experiments. The microfluidic system is endothelialized, and the regional tissue injury-induced thrombosis is then realized through optical irradiation. Whole blood is perfused with MPs, and the invasion process visualized and recorded. The mouse model shows a cumulative risk in the blood with continuous exposure to MPs (P-value < 0.0001). The on-chip results show that MP invasion leads to decreased binding of fibrin to platelets (P-value < 0.0001), which is consistent with the results of the in vitro experiments, and shows a high risk of thrombus shedding in real blood flow compared with normal thrombus. This work provides a new method to further reveal MP-related health risks.

Revealing Fine Structures of the Retinal Receptive Field by Deep-Learning Networks.

Deep convolutional neural networks (CNNs) have demonstrated impressive performance on many visual tasks. Recently, they became useful models for the visual system in neuroscience. However, it is still not clear what is learned by CNNs in terms of neuronal circuits. When a deep CNN with many layers is used for the visual system, it is not easy to compare the structure components of CNNs with possible neuroscience underpinnings due to highly complex circuits from the retina to the higher visual cortex. Here, we address this issue by focusing on single retinal ganglion cells with biophysical models and recording data from animals. By training CNNs with white noise images to predict neuronal responses, we found that fine structures of the retinal receptive field can be revealed. Specifically, convolutional filters learned are resembling biological components of the retinal circuit. This suggests that a CNN learning from one single retinal cell reveals a minimal neural network carried out in this cell. Furthermore, when CNNs learned from different cells are transferred between cells, there is a diversity of transfer learning performance, which indicates that CNNs are cell specific. Moreover, when CNNs are transferred between different types of input images, here white noise versus natural images, transfer learning shows a good performance, which implies that CNNs indeed capture the full computational ability of a single retinal cell for different inputs. Taken together, these results suggest that CNNs could be used to reveal structure components of neuronal circuits, and provide a powerful model for neural system identification.

Spike-based Motion Estimation for Object Tracking through Bio-inspired Unsupervised Learning.

Neuromorphic vision sensors, whose pixels output events/spikes asynchronously with a high temporal resolution according to the scene radiance change, are naturally appropriate for capturing high-speed motion in the scenes. However, how to utilize the events/spikes to smoothly track high-speed moving objects is still a challenging problem. Existing approaches either employ time-consuming iterative optimization, or require large amounts of labeled data to train the object detector. To this end, we propose a bio-inspired unsupervised learning framework, which takes advantage of the spatiotemporal information of events/spikes generated by neuromorphic vision sensors to capture the intrinsic motion patterns. Without off-line training, our models can filter the redundant signals with dynamic adaption module based on short-term plasticity, and extract the motion patterns with motion estimation module based on the spike-timing-dependent plasticity. Combined with the spatiotemporal and motion information of the filtered spike stream, the traditional DBSCAN clustering algorithm and Kalman filter can effectively track multiple targets in extreme scenes. We evaluate the proposed unsupervised framework for object detection and tracking tasks on synthetic data, publicly available event-based datasets, and spiking camera datasets. The experiment results show that the proposed model can robustly detect and smoothly track the moving targets on various challenging scenarios and outperforms state-of-the-art approaches.