Climate-related naturally occurring epimutation and their roles in plant adaptation in A. thaliana.

DNA methylation has been proposed to be an important mechanism that allows plants to respond to their environments sometimes entirely uncoupled from genetic variation. To understand the genetic basis, biological functions and climatic relationships of DNA methylation at a population scale in Arabidopsis thaliana, we performed a genome-wide association analysis with high-quality single nucleotide polymorphisms (SNPs), and found that ~56% on average, especially in the CHH sequence context (71%), of the differentially methylated regions (DMRs) are not tagged by SNPs. Among them, a total of 3235 DMRs are significantly associated with gene expressions and potentially heritable. 655 of the 3235 DMRs are associated with climatic variables, and we experimentally verified one of them, HEI10 (HUMAN ENHANCER OF CELL INVASION NO.10). Such epigenetic loci could be subjected to natural selection thereby affecting plant adaptation, and would be expected to be an indicator of accessions at risk. We therefore incorporated these climate-related DMRs into a gradient forest model, and found that the natural A. thaliana accessions in Southern Europe that may be most at risk under future climate change. Our findings highlight the importance of integrating DNA methylation that is independent of genetic variations, and climatic data to predict plants' vulnerability to future climate change.

Buyang Huanwu decoction promotes angiogenesis after cerebral ischemia through modulating caveolin-1-mediated exosome MALAT1/YAP1/HIF-1α axis.

BACKGROUND: Angiogenesis is an effective method for promoting neurological function recovery after cerebral ischemia (CI). Buyang Huanwu decoction (BHD) is a traditional Chinese medicinal recipe that is frequently employed for CI treatment. Previous investigations have validated that it promotes angiogenesis following CI. Nevertheless, the precise mechanism by which it does this has yet to be completely understood. OBJECTIVE: This study aims to examine the underlying mechanism through which BHD facilitates angiogenesis following CI by regulating the exosomal MALAT1/YAP1/HIF-1α signaling axis, specifically via the involvement of caveolin-1 (Cav1), an endocytosis-associated protein. METHODS: A CI model was created using middle cerebral artery occlusion (MCAO). Following the administration of multiple doses of BHD, various parameters, including the neurobehavioral score, pathological damage, and angiogenesis, were assessed in each group of mice to identify the optimal dosage of BHD for treating CI. The molecular processes underlying the angiogenic implications of BHD following CI were investigated exhaustively by employing single-cell sequencing. Finally, the involvement of Cav1 was confirmed in Cav1 knockout mice and Cav1-silenced stably transfected strains to validate the mechanism by which BHD increases angiogenesis following CI. RESULTS: BHD could promote angiogenesis after CI. Single-cell sequencing results suggested that its potential mechanism of action might be connected with Cav1 and the exosomal MALAT1/YAP1/HIF-1α signaling axis. BHD could promote angiogenesis after CI by regulating the exosomal MALAT1/YAP1/HIF-1α axis through Cav1, as validated in vivo and in vitro experiments. Accordingly, Cav1 may be a key target of BHD in promoting angiogenesis after CI. CONCLUSION: This investigation represents the initial attempt to comprehensively ascertain the underlying mechanism of action of BHD in treating CI using single-cell sequencing, gene-knockout mice, and stable transfected cell lines, potentially associated with the modulation of the exosomal MALAT1/YAP1/HIF-1α axis by Cav1. Our findings offer novel empirical evidence for unraveling the regulatory pathways through which Cav1 participates in angiogenesis following CI and shed light on the potential mechanisms of BHD.

Structural deep multi-view clustering with integrated abstraction and detail.

Deep multi-view clustering, which can obtain complementary information from different views, has received considerable attention in recent years. Although some efforts have been made and achieve decent performances, most of them overlook the structural information and are susceptible to poor quality views, which may seriously restrict the capacity for clustering. To this end, we propose Structural deep Multi-View Clustering with integrated abstraction and detail (SMVC). Specifically, multi-layer perceptrons are used to extract features from specific views, which are then concatenated to form the global features. Besides, a global target distribution is constructed and guides the soft cluster assignments of specific views. In addition to the exploitation of the top-level abstraction, we also design the mining of the underlying details. We construct instance-level contrastive learning using high-order adjacency matrices, which has an equivalent effect to graph attention network and reduces feature redundancy. By integrating the top-level abstraction and underlying detail into a unified framework, our model can jointly optimize the cluster assignments and feature embeddings. Extensive experiments on four benchmark datasets have demonstrated that the proposed SMVC consistently outperforms the state-of-the-art methods.

Magnetoresistance properties in nickel-catalyzed, air-stable, uniform, and transfer-free graphene.

A transfer-free graphene with high magnetoresistance (MR) and air stability has been synthesized using nickel-catalyzed atmospheric pressure chemical vapor deposition. The Raman spectrum and Raman mapping reveal the monolayer structure of the transfer-free graphene, which has low defect density, high uniformity, and high coverage (>90%). The temperature-dependent (from 5 to 300 K) current-voltage (I-V) and resistance measurements are performed, showing the semiconductor properties of the transfer-free graphene. Moreover, the MR of the transfer-free graphene has been measured over a wide temperature range (5-300 K) under a magnetic field of 0 to 1 T. As a result of the Lorentz force dominating above 30 K, the transfer-free graphene exhibits positive MR values, reaching ∼8.7% at 300 K under a magnetic field (1 Tesla). On the other hand, MR values are negative below 30 K due to the predominance of the weak localization effect. Furthermore, the temperature-dependent MR values of transfer-free graphene are almost identical with and without a vacuum annealing process, indicating that there are low density of defects and impurities after graphene fabrication processes so as to apply in air-stable sensor applications. This study opens avenues to develop 2D nanomaterial-based sensors for commercial applications in future devices.

Integrating network pharmacology and bioinformatics to explore and experimentally verify the regulatory effect of Buyang Huanwu decoction on glycolysis and angiogenesis after cerebral infarction.

ETHNOPHARMACOLOGICAL RELEVANCE: Promoting the recovery of cerebral blood circulation after cerebral infarction (CI) is an important intervention. Buyang Huanwu decoction (BHD) is a classic prescription for treating CI that promotes angiogenesis. Cytoplasmic glycolysis ischaemic-region cells after CI may be highly activated to maintain metabolic activity under hypoxia. From the perspective of long-term maintenance of glycolytic metabolism in the ischaemic area after CI, it may be beneficial to promote angiogenesis and maintain glial cell activation and neuronal survival. In this context, the regulatory relationship of lncRNAs and miRNAs with mRNAs is worthy of attention. Mining the competitive binding relationships among RNAs will aid in the screening of key gene targets post-CI. In this study, network pharmacology and bioinformatics were used to construct a ceRNA network, screen key targets, and explore the effect of glycolysis on angiogenesis during BHD-mediated CI regulation. AIM OF THE STUDY: This study aimed to explore the effect of BHD on angiogenesis after glycolysis regulation in CI. MATERIALS AND METHODS: According to the 21 active BHD ingredients we identified by our research team, we conducted network pharmacology. BHD targets that can regulate glycolysis and angiogenesis after CI were screened from the GeneCards, CTD and OMIM databases. We retrieved CI-related datasets from the GEO database and screened for differentially expressed lncRNAs and miRNAs. LncRNA‒miRNA-mRNA/TF targeting relationships were screened and organized with the miRcode, miRDB, TargetScan, miRWalk, and TransmiR v2.0 databases. Cytoscape was used to construct an lncRNA‒miRNA-mRNA/TF ceRNA network. Through BioGPS, key mRNAs/TFs in the network were screened for enrichment analysis. Animal experiments were then conducted to validate some key mRNAs/TFs and enriched signalling pathways. RESULTS: PFKFB3 and other genes may help regulate glycolysis and angiogenesis through AMPK and other signalling pathways. The anti-CI effect of BHD may involve maintaining activation of genes such as AMPK and PFKFB3 in the ischaemic cortex, maintaining moderate glycolysis levels in brain tissue, and promoting angiogenesis. CONCLUSION: BHD can regulate glycolysis and promote angiogenesis after CI through multiple pathways and targets, in which AMPK signalling pathway activation may be important.

Measurement of the 5S1/2 to 5D5/2 two-photon clock transition frequency of rubidium-85 in high vacuum.

We present a scheme to precisely resolve the unperturbed line shape of an optical rubidium clock transition in a high vacuum, by which we avoided the systematic errors of "collision shift" and "modulation shift." The spectral resolution resolved by this scheme is significantly improved such that we can use "Zeeman broadening" to inspect the stray magnetic field, through which we were able to compensate the magnetic field inside the Rb cells to be below 10-3 Gauss. We thus update the absolute frequency of the clock transition and propose a standard operation procedure (SOP) for the clock self-calibration.

In situ rapid synthesis of ionic liquid/ionic covalent organic framework composites for CO2 fixation.

IL/ICOF composites were in situ synthesized via a one-pot route in half an hour under ambient conditions for catalytic cycloaddition of CO2 with epoxides into cyclic carbonates. The prepared composites feature a decent CO2 adsorption capacity of 1.63 mmol g-1 at 273 K and 1 bar and exhibit excellent catalytic performance in terms of yield and durability. This work may pave a new way to design and construct functionalized porous organic frameworks as heterogeneous catalysts for CO2 capture and conversion.

Construction of amidoxime-functionalized magnetic hydroxyapatite with enhanced uranium extraction performance from aqueous solution and seawater.

A novel amidoxime-functionalized magnetic hydroxyapatite (AFNH) was successfuly fabricated to extract uranium from aqueous solution and seawater. The introduction of amidoxime group not only increased the number of active site of AFNH to speed up the adsorption rate and increase the extraction capacity, but also adjusted the optimal extraction pH from 4 to 8, which was beneficial for capturing uranium from seawater. The maximum adsorption capacity and adsorption efficiency at pH 8 were 945.2 mg g-1 and 99.2%, respectively. AFNH still had good removal efficiency (above 90%) after five cycles, indicating the good regeneration of AFNH. After uranium adsorption, AFNH could be easily recycled by magnetic separation due to its magnetism. In simulated seawater, AFNH also showed excellent uranium removal performance with high adsorption efficiency (84.9%) and adsorption capacity (1.70 mg g-1). Furthermore, the 14-day uranium extraction capacity of AFNH in natural seawater could reach 5.93 mg g-1. The SEM, FTIR, XRD and XPS analyses showed that the enhanced uranium extraction performance of AFNH was mainly attributed to electrostatic interaction, complexation and co-precipitation. In conclusion, AFNH was expected to be a candidate as adsorbent with great potential in extracting uranium from seawater.

Structure-aware deep clustering network based on contrastive learning.

Recently, deep clustering has been extensively employed for various data mining tasks, and it can be divided into auto-encoder (AE)-based and graph neural networks (GNN)-based methods. However, existing AE-based methods fall short in effectively extracting structural information, while GNN suffer from smoothing and heterophily. Although methods that combine AE and GNN achieve impressive performance, there remains an inadequate balance between preserving the raw structure and exploring the underlying structure. Accordingly, we propose a novel network named Structure-Aware Deep Clustering network (SADC). Firstly, we compute the cumulative influence of non-adjacent nodes at multiple depths and, thus, enhance the adjacency matrix. Secondly, an enhanced graph auto-encoder is designed. Thirdly, the latent space of AE is endowed with the ability to perceive the raw structure during the learning process. Besides, we design self-supervised mechanisms to achieve co-optimization of node representation learning and topology learning. A new loss function is designed to preserve the inherent structure while also allowing for exploration of latent data structure. Extensive experiments on six benchmark datasets validate that our method outperforms state-of-the-art methods.

Enhancing Prediction of Forelimb Movement Trajectory through a Calibrating-Feedback Paradigm Incorporating RAT Primary Motor and Agranular Cortical Ensemble Activity in the Goal-Directed Reaching Task.

Complete reaching movements involve target sensing, motor planning, and arm movement execution, and this process requires the integration and communication of various brain regions. Previously, reaching movements have been decoded successfully from the motor cortex (M1) and applied to prosthetic control. However, most studies attempted to decode neural activities from a single brain region, resulting in reduced decoding accuracy during visually guided reaching motions. To enhance the decoding accuracy of visually guided forelimb reaching movements, we propose a parallel computing neural network using both M1 and medial agranular cortex (AGm) neural activities of rats to predict forelimb-reaching movements. The proposed network decodes M1 neural activities into the primary components of the forelimb movement and decodes AGm neural activities into internal feedforward information to calibrate the forelimb movement in a goal-reaching movement. We demonstrate that using AGm neural activity to calibrate M1 predicted forelimb movement can improve decoding performance significantly compared to neural decoders without calibration. We also show that the M1 and AGm neural activities contribute to controlling forelimb movement during goal-reaching movements, and we report an increase in the power of the local field potential (LFP) in beta and gamma bands over AGm in response to a change in the target distance, which may involve sensorimotor transformation and communication between the visual cortex and AGm when preparing for an upcoming reaching movement. The proposed parallel computing neural network with the internal feedback model improves prediction accuracy for goal-reaching movements.

Efficient separation of uranium(VI) from aqueous solution using magnetic Co/Al layered double oxides coated with carbon dots.

Herein, magnetic layered double oxides coated with carbon dots (MLCs) were synthesized through introducing sodium dodecylbenzene sulfonate and FeCl2 into Co/Al LDH for capturing uranium from aqueous solution. When the molar ratio of Co to Al was 4 : 1, the MLC composite possessed the strongest affinity to uranium(VI) in solution with short equilibrium time (<160 min), high adsorption efficiency (94.31%) and large removal capacity (513.85 mg g-1). The adsorption behavior of MLCs for uranium(VI) was well fitted with Langmuir and pseudo-second-order models, suggesting that the monolayer chemical adsorption was the rate-limiting step. Besides, MLC-3 could be reused by using 0.15 mol L-1 ethylene diamine tetraacetic acid as an eluent and the removal percentage still remained at a high level (>83.3%) after 5 adsorption/desorption cycles. Redox reaction, chemical complexation and electrostatic attraction were proved to play significant roles in uranium(VI) separation. Therefore, MLC-3 could be used as a potential adsorbent in uranium(VI)-containing wastewater treatment due to its excellent adsorption performance for uranium(VI).

Assessment of coastal sustainable development along the maritime silk road using an integrated natural-economic-social (NES) ecosystem.

Understanding spatial change and its driving factors behind coastal development is essential for coastal management and restoration. There is an urgent need for quantitative assessments of sustainable development in the coastal ecosystems that are most affected by anthropogenic activities and climate change. This study built a theme-based evaluation methodology with the Natural-Economic-Social (NES) complex ecosystem and proposed an evaluation system of coastal sustainable development (CSD) to understand the complex interactions between coastal ecosystems and anthropogenic activities. The approach revealed the levels of coastal natural, economic, and social sustainable development in the countries along the Maritime Silk Road (MSR) from 2010 to 2020. The results showed (1) a decreasing trend for coastal sustainable development between 2010 and 2015 and a rapid increasing trend between 2015 and 2020; (2) spatially varied CSD, with higher levels in Europe and Southeast Asia and lower levels in South and West Asia and North Africa; and (3) a strong influence on CSD by a combination of economic and social factors and relatively little influence by natural factors. The study further assessed the natural, economic, and social development scores for 41 countries and compared them with the mean scores (MSR) to classify coastal development patterns into three stages (favorable, transitional, and unfavorable). Finally, in the context of the 2030 Agenda for Sustainable Development, the study highlighted the importance of more refined global indicators for CSD assessments.

Buyang Huanwu Decoction alleviates cerebral ischemic injury through modulating caveolin-1-mediated mitochondrial quality control.

Introduction: Mitochondrial quality control (MQC) is an important mechanism of neural repair after cerebral ischemia (CI). Recent studies have shown that caveolin-1 (Cav-1) is an important signaling molecule in the process of CI injury, but its mechanism of regulating MQC after CI is still unclear. Buyang Huanwu Decoction (BHD) is a classic traditional Chinese medicine formula that is often used to treat CI. Unfortunately, its mechanism of action is still obscure. Methods: In this study, we tested the hypothesis that BHD can regulate MQC through Cav-1 and exert an anti-cerebral ischemia injury effect. We used Cav-1 knockout mice and their homologous wild-type mice, replicated middle cerebral artery occlusion (MCAO) model and BHD intervention. Neurobehavioral scores and pathological detection were used to evaluate neurological function and neuron damage, transmission electron microscopy and enzymology detection of mitochondrial damage. Finally, western blot and RT-qPCR expression of MQC-related molecules were tested. Results: After CI, mice showed neurologic impairment, neuronal damage, and significant destruction of mitochondrial morphology and function, and MQC was imbalanced. Cav-1 deletion aggravated the damage to neurological function, neurons, mitochondrial morphology and mitochondrial function after CI, aggravated the imbalance of mitochondrial dynamics, and inhibited mitophagy and biosynthesis. BHD can maintain MQC homeostasis after CI through Cav-1 and improve CI injury. Discussion: Cav-1 can affect CI injury by regulating MQC, and this mechanism may be another target of BHD for anti-cerebral ischemia injury.

Fabrication of phosphoric-crosslinked chitosan@g-C3N4 gel beads for uranium(VI) separation from aqueous solution.

In this work, a novel g-C3N4 filled, phosphoric-crosslinked chitosan gel bead (P-CS@CN) was successfully prepared to adsorb U(VI) from water. The separation performance of chitosan was improved by introducing more functional groups. At pH 5 and 298 K, the adsorption efficiency and adsorption capacity could reach 98.0 % and 416.7 mg g-1, respectively. After adsorption, the morphological structure of P-CS@CN did not change and adsorption efficiency remained above 90 % after 5 cycles. P-CS@CN exhibited an excellent applicability in water environment based on dynamic adsorption experiments. Thermodynamic analyses demonstrated the value of ΔG, manifesting the spontaneity of U(VI) adsorption process on P-CS@CN. The positive values of ΔH and ΔS showed that the U(VI) removal behavior of P-CS@CN was an endothermic reaction, indicating that the increase of temperature was great benefit to the removal. The adsorption mechanism of P-CS@CN gel bead could be summarized as the complexation reaction with the surface functional groups. This study not only developed an efficient adsorbent for the treatment of radioactive pollutants, but also provided a simple and feasible strategy for the modification of chitosan-based adsorption materials.

High-resolution Cs-Rb two-photon spectrometer for directly stabilizing a Ti:sapphire comb laser.

In this paper, we present a simple scheme for efficiently removing the residual Doppler background of a comb laser based two-photon spectrometer to be better than 10-3 background-to-signal ratio. We applied this scheme to stabilize the frequencies of a mode-locked Ti:sapphire laser directly referring to the cesium 6S-8S transition and rubidium 5S-5D transition. We suggest a standard operation procedure (SOP) for the fully direct comb laser stabilization and evaluate the frequency of two spectral lines at a certain temperature, by which we demonstrate an all-atomic-transition-based Ti:sapphire comb laser merely via a 6-cm glass cell.

Construction and function of a high-efficient synthetic bacterial consortium to degrade aromatic VOCs.

Aromatic volatile organic compounds (VOCs) are a type of common pollution form in chemical contaminated sites. In this study, seven aromatic VOCs such as benzene, toluene, ethylbenzene, chlorobenzene, m-xylene, p-chlorotoluene and p-chlorotrifluorotoluene were used as the only carbon source, and four strains of highly efficient degrading bacteria were screened from the soil of chemical contaminated sites, then the synthetic bacterial consortium was constructed after mixing with an existing functional strain (Bacillus benzoevorans) preserved in the laboratory. After that, the synthetic bacterial consortium was used to explore the degradation effect of simulated aromatic VOCs polluted wastewater. The results showed that the functional bacterium could metabolize with aromatic VOCs as the only carbon source and energy. Meanwhile, the growth of the synthetic bacterial consortium increased with the additional carbon resources and the alternative of organic nitrogen source. Ultimately, the applicability of the synthetic bacterial consortium in organic contaminated sites was explored through the study of broad-spectrum activity.

A UPLC-Q-TOF/MS and network pharmacology method to explore the mechanism of Anhua fuzhuan tea intervention in nonalcoholic fatty liver disease.

The possible mechanism by which the active components of Anhua fuzhuan tea act on FAM in NAFLD lesions was investigated. 83 components of Anhua fuzhuan tea were analysed by UPLC-Q-TOF/MS. Luteolin-7-rutinoside and other compounds were first discovered in fuzhuan tea. According to the TCMSP database and the Molinspiration website tool to predict and review the literature reports, 78 compounds were identified in fuzhuan tea with possible biological activities. The PharmMapper, Swiss target prediction, and SuperPred databases were used to predict the action targets of biologically active compounds. The GeneCards, CTD, and OMIM databases were mined for NAFLD and FAM genes. Then, a fuzhuan Tea-NAFLD-FAM Venn diagram was constructed. Using the STRING database and CytoHubba program of Cytoscape software, protein interaction analysis was performed, and 16 key genes, including PPARG, were screened. GO function and KEGG enrichment analyses of the screened key genes showed that Anhua fuzhuan tea may regulate FAM in the process of NAFLD through the AMPK signalling pathway, nonalcoholic fatty liver disease pathway, etc. After constructing an active ingredient-key target-pathway map with Cytoscape software, combined with literature reports and BioGPS database analysis, we believe that among the 16 key genes, SREBF1, FASN, ACADM, HMGCR, and FABP1 have potential in the treatment of NAFLD. Animal experiments confirmed the effect of Anhua fuzhuan tea in improving NAFLD and confirmed that this tea can interfere with the gene expression of the above five targets by the AMPK/PPAR pathway, providing support for Anhua fuzhuan tea interfering with FAM in NAFLD lesions.

Proof of Concept for Sustainable Manufacturing of Neural Electrode Array for In Vivo Recording.

Increasing requirements for neural implantation are helping to expand our understanding of nervous systems and generate new developmental approaches. It is thanks to advanced semiconductor technologies that we can achieve the high-density complementary metal-oxide-semiconductor electrode array for the improvement of the quantity and quality of neural recordings. Although the microfabricated neural implantable device holds much promise in the biosensing field, there are some significant technological challenges. The most advanced neural implantable device relies on complex semiconductor manufacturing processes, which are required for the use of expensive masks and specific clean room facilities. In addition, these processes based on a conventional photolithography technique are suitable for mass production, which is not applicable for custom-made manufacturing in response to individual experimental requirements. The microfabricated complexity of the implantable neural device is increasing, as is the associated energy consumption, and corresponding emissions of carbon dioxide and other greenhouse gases, resulting in environmental deterioration. Herein, we developed a fabless fabricated process for a neural electrode array that was simple, fast, sustainable, and customizable. An effective strategy to produce conductive patterns as the redistribution layers (RDLs) includes implementing microelectrodes, traces, and bonding pads onto the polyimide (PI) substrate by laser micromachining techniques combined with the drop coating of the silver glue to stack the laser grooving lines. The process of electroplating platinum on the RDLs was performed to increase corresponding conductivity. Sequentially, Parylene C was deposited onto the PI substrate to form the insulation layer for the protection of inner RDLs. Following the deposition of Parylene C, the via holes over microelectrodes and the corresponding probe shape of the neural electrode array was also etched by laser micromachining. To increase the neural recording capability, three-dimensional microelectrodes with a high surface area were formed by electroplating gold. Our eco-electrode array showed reliable electrical characteristics of impedance under harsh cyclic bending conditions of over 90 degrees. For in vivo application, our flexible neural electrode array demonstrated more stable and higher neural recording quality and better biocompatibility as well during the 2-week implantation compared with those of the silicon-based neural electrode array. In this study, our proposed eco-manufacturing process for fabricating the neural electrode array reduced 63 times of carbon emissions compared to the traditional semiconductor manufacturing process and provided freedom in the customized design of the implantable electronic devices as well.

Mapping seagrasses on the basis of Sentinel-2 images under tidal change.

Tidal variations make the water bodies in satellite remote sensing images on different shooting dates have different inundation ranges and depths. Although the underwater substrates do not change, the spectral properties differ due to attenuation effects. These differences have an impact on the results when multi-temporal remote sensing images are used to analyze seagrasses. This paper proposes a remote sensing mapping method for seagrasses taking the tidal influence, using the seagrasses growth area in Xincun Bay, Hainan Province, China as a case study. a) The seagrasses growth area was determined from remote sensing images. The seagrasses were divided into two types: the seagrasses exposed to water surface or tidal flats (non-submerged seagrasses) and the seagrasses submerged in water (submerged seagrasses). b) The spectral features of seagrasses in Sentienl-2 image were analyzed. We found that the spectral characteristics of non-submerged seagrasses were similar to terrestrial vegetation and these seagrasses could be extracted by using NDVI. The submerged seagrasses spectral was different, forming a reflection peak at the first vegetation red edge band (i.e.705 nm) in Sentinel-2 images. This reflection peak was used to design the Submerged Seagrasses Identification Index (SSII) for extracting underwater seagrass. c) The extraction results of non-submerged seagrasses and submerged seagrasses were merged to map the seagrasses in the study area. The experimental results show that the mapping method proposed in this study can fully consider the influence of tidal changes in remote sensing images on seagrasses identification. The SSII constructed based on Sentinel-2 images extracted submerged seagrasses effectively. This study will provide references to remote sensing mapping of seagrasses and integrated ecological management in coastal zones.

Convolutional neural network for detecting rib fractures on chest radiographs: a feasibility study.

BACKGROUND: Chest radiography is the standard investigation for identifying rib fractures. The application of artificial intelligence (AI) for detecting rib fractures on chest radiographs is limited by image quality control and multilesion screening. To our knowledge, few studies have developed and verified the performance of an AI model for detecting rib fractures by using multi-center radiographs. And existing studies using chest radiographs for multiple rib fracture detection have used more complex and slower detection algorithms, so we aimed to create a multiple rib fracture detection model by using a convolutional neural network (CNN), based on multi-center and quality-normalised chest radiographs. METHODS: A total of 1080 radiographs with rib fractures were obtained and randomly divided into the training set (918 radiographs, 85%) and the testing set (162 radiographs, 15%). An object detection CNN, You Only Look Once v3 (YOLOv3), was adopted to build the detection model. Receiver operating characteristic (ROC) and free-response ROC (FROC) were used to evaluate the model's performance. A joint testing group of 162 radiographs with rib fractures and 233 radiographs without rib fractures was used as the internal testing set. Furthermore, an additional 201 radiographs, 121 with rib fractures and 80 without rib fractures, were independently validated to compare the CNN model performance with the diagnostic efficiency of radiologists. RESULTS: The sensitivity of the model in the training and testing sets was 92.0% and 91.1%, respectively, and the precision was 68.0% and 81.6%, respectively. FROC in the testing set showed that the sensitivity for whole-lesion detection reached 91.3% when the false-positive of each case was 0.56. In the joint testing group, the case-level accuracy, sensitivity, specificity, and area under the curve were 85.1%, 93.2%, 79.4%, and 0.92, respectively. At the fracture level and the case level in the independent validation set, the accuracy and sensitivity of the CNN model were always higher or close to radiologists' readings. CONCLUSIONS: The CNN model, based on YOLOv3, was sensitive for detecting rib fractures on chest radiographs and showed great potential in the preliminary screening of rib fractures, which indicated that CNN can help reduce missed diagnoses and relieve radiologists' workload. In this study, we developed and verified the performance of a novel CNN model for rib fracture detection by using radiography.