Orthogonal Network Research Articles

Fuzzy Attention-Based Border Rendering Orthogonal Network for Lung Organ Segmentation

Fuzzy Attention-Based Border Rendering Orthogonal Network for Lung Organ Segmentation

In Situ UNIversal Orthogonal Network (UNION) Bioink Deposition for Direct Delivery of Corneal Stromal Stem Cells to Corneal Wounds.

The scarcity of human donor corneal graft tissue worldwide available for corneal transplantation necessitates the development of alternative therapeutic strategies for treating patients with corneal blindness. Corneal stromal stem cells (CSSCs) have the potential to address this global shortage by allowing a single donor cornea to treat multiple patients. To directly deliver CSSCs to corneal defects within an engineered biomatrix, we developed a UNIversal Orthogonal Network (UNION) collagen bioink that crosslinks in situ with a bioorthogonal, covalent chemistry. This cell-gel therapy is optically transparent, stable against contraction forces exerted by CSSCs, and permissive to the efficient growth of corneal epithelial cells. Furthermore, CSSCs remain viable within the UNION collagen gel precursor solution under standard storage and transportation conditions. This approach promoted corneal transparency and re-epithelialization in a rabbit anterior lamellar keratoplasty model, indicating that the UNION collagen bioink serves effectively as an in situ -forming, suture-free therapy for delivering CSSCs to corneal wounds. TEASER. Corneal stem cells are delivered within chemically crosslinked collagen as a transparent, regenerative biomaterial therapy.

Efficient production of pullulan by Aureobasidium pullulans using a multi-objective optimization strategy with orthogonal experimental design coupling artificial neural network and genetic algorithm

Efficient pullulan production has long been a central research focus. This study used maltodextrin as the carbon source for pullulan production by Aureobasidium pullulans fermentation. A hybrid optimization approach, integrating orthogonal experimental design (OED), backpropagation artificial neural network (BP-ANN), and elite strategy non-dominated sequential genetic algorithm-II (NSGA-II), was developed. Range analysis based on OED revealed that MgSO4·7H2O significantly affects production but less impacts molecular weight, while pH notably influences molecular weight with a lesser effect on production, underscoring the need for multi-objective optimization. The BP-ANN model showed strong predictive capabilities, with goodness-of-fit values of 0.984 and 0.980 for production and molecular weight, respectively. Using this model as the fitness function for the optimization algorithm enhanced efficiency. Taking cost factors into account, the BP-ANN-NSGA-II algorithm identified the optimal fermentation medium conditions, resulting in a 6.89 % increase in production, a 368.97 % increase in molecular weight, and a 42.49 % reduction in cost. The maximum comprehensive optimization efficiency is 63.73 %, and the multi-objective optimization is better than the single objective optimization. This method significantly guides the improvement of pullulan fermentation optimization efficiency.

The hinge-1 domain of Flna is not necessary for diverse physiological functions in mice.

The filamins are cytoskeletal binding proteins that dynamically crosslink actin into orthogonal networks or bundle it into stress fibres. The domain structure of filamin proteins is very well characterised, with an N-terminal actin-binding region, followed by 24 immunoglobulin-like repeat units. The repeat domains are separated into distinct segments by two regions of low-complexity known as hinge-1 and hinge-2. The role of hinge-1 especially has been proposed to be essential for protein function as it provides flexibility to the otherwise rigid protein, and is a target for cleavage by calpain. Hinge-1 protects cells from otherwise destructive forces, and the products of calpain cleavage are involved in critical cellular signalling processes, such as survival during hypoxia. Pathogenic variants in FLNA encoding Filamin A, including those that remove the hinge-1 domain, cause a wide range of survivable developmental disorders. In contrast, complete loss of function of this gene is embryonic lethal in human and mouse. In this study, we show that removing filamin A hinge-1 from mouse (FlnaΔH1), while preserving its expression level leads to no obvious developmental phenotype. Detailed characterisation of the skeletons of FlnaΔH1 mice showed no skeletal phenotype reminiscent of that found in the FLNA-causing skeletal dysplasia. Furthermore, nuclear functions of FLNA are maintained with loss of Filamin A hinge-1. We conclude that hinge-1 is dispensable for filamin A protein function during development over the murine lifespan.

Integrative in vivo analysis of the ethanolamine utilization bacterial microcompartment in Escherichia coli.

Bacterial microcompartments (BMCs) are self-assembling protein megacomplexes that encapsulate metabolic pathways. Although approximately 20% of sequenced bacterial genomes contain operons encoding putative BMCs, few have been thoroughly characterized, nor any in the most studied Escherichia coli strains. We used an interdisciplinary approach to gain deep molecular and functional insights into the ethanolamine utilization (Eut) BMC system encoded by the eut operon in E. coli K-12. The eut genotype was linked with the ethanolamine utilization phenotype using deletion and overexpression mutants. The subcellular dynamics and morphology of the E. coli Eut BMCs were characterized in cellula by fluorescence microscopy and electron (cryo)microscopy. The minimal proteome reorganization required for ethanolamine utilization and the in vivo stoichiometric composition of the Eut BMC were determined by quantitative proteomics. Finally, the first flux map connecting the Eut BMC with central metabolism in cellula was obtained by genome-scale modeling and 13C-fluxomics. Our results reveal that contrary to previous suggestions, ethanolamine serves both as a nitrogen and a carbon source in E. coli K-12, while also contributing to significant metabolic overflow. Overall, this study provides a quantitative molecular and functional understanding of the BMCs involved in ethanolamine assimilation by E. coli.IMPORTANCEThe properties of bacterial microcompartments make them an ideal tool for building orthogonal network structures with minimal interactions with native metabolic and regulatory networks. However, this requires an understanding of how BMCs work natively. In this study, we combined genetic manipulation, multi-omics, modeling, and microscopy to address this issue for Eut BMCs. We show that the Eut BMC in Escherichia coli turns ethanolamine into usable carbon and nitrogen substrates to sustain growth. These results improve our understanding of compartmentalization in a widely used bacterial chassis.

Independent Configuration and Reprogramming of Porous Characters in Macroporous Hydrogel Enabled by the Orthogonal Dynamic Network.

Macroporous hydrogels have attracted much attention in both industry and academia, where the morphological characteristics of pores are essential. Despite significant improvements on regulating porous structures, the independent configuration and reprogramming of porosity and pore size still remain challenging owing to the lack of a chemical design to decouple the mechanism for adjusting each characteristic. Here, we report a strategy to adaptively control porous features relying on an orthogonal dynamic network. Disulfide bonds are employed to relax polymer chains during freezing via UV irradiation, thus, generating pores in hydrogels. On such a basis, the porosity is continuously switched from 0 to 75% by controlling network relaxation ratios. Subsequently, the pore size is further reversibly manipulated through the association or dissociation of dynamic metallic coordination. As a result, the porosity and pore size achieved independent configurations. Meanwhile, the dynamic nature of the network makes it possible to reprogram the porous character of a prepared hydrogel. Beyond these, the photopatterning of pores represents the capability to regulate the third feature. Our strategy provides an effective way to arbitrarily manipulate porous morphologies, which can inspire the design of future functional porous materials.

Secure MIMO communication in energy harvesting-assisted NOMA Cognitive Radio Network with jamming under hardware impairment

Energy harvesting (EH)-assisted non-orthogonal multiple access (NOMA) cognitive radio (CR) networks allow simultaneous transmission of multiple secondary user signals on primary frequency bands with harvested energy, enhancing spectral, spectrum utilization, and energy efficiencies. Although multiple antennas are used for efficient energy transfer and signal transceiving, multiple-input multiple-output (MIMO) communication in these networks is facing reliability/security performance degradation due to nonlinear EH, hardware impairment (HWi), and wire-tapping. The paper aims to numerically evaluate the security and reliability of MIMO communication in EH-assisted NOMA CR networks with jamming (MehNOwJ) under such effects. The results indicate that MehNOwJ prevents full outage and achieves optimum performance with proper parameter selection of preset spectral efficiency, power saturation threshold, EH duration, number of antennas of jammer. In addition, the performance improves with an accreting quantity of antennas but experiences saturation. Moreover, MehNOwJ drastically outperforms alternative approaches (MIMO communication in EH-assisted orthogonal multiple access CR network with jamming and MIMO communication in EH-assisted NOMA CR network without jamming), offering insights into the benefits of combining NOMA and jamming techniques.

Charting the importance of filamin A posttranslational modifications.

Filamin A is an essential protein in the cell cytoskeleton because of its actin binding properties and unique homodimer rod-shaped structure, which organises actin into three-dimensional orthogonal networks imperative to cell motility, spreading and adhesion. Filamin A is subject to extensive posttranslational modification (PTM) which serves to co-ordinate cellular architecture and to modulate its large protein-protein interaction network which is key to the protein's role as a cellular signalling hub. Characterised PTMs include phosphorylation, irreversible cleavage, ubiquitin mediated degradation, hydroxylation and O-GlcNAcylation, with preliminary evidence of tyrosylation, carbonylation and acetylation. Each modification and its relation to filamin A function will be described here. These modifications are often aberrantly applied in a range of diseases including, but not limited to, cancer, cardiovascular disease and neurological disease and we discuss the concept of target specific PTMs with novel therapeutic modalities. In summary, our review represents a topical 'one-stop-shop' that enables understanding of filamin A function in cell homeostasis and provides insight into how a variety of modifications add an extra level of Filamin A control.

Sedimentary basins, hydrocarbons, graphite, coal, and Cu-Au deposits - from Mongolia to the Pacific margin: Interplay between the ubiquitous orthogonal fracture network and Global Wrench Tectonics

Mongolia is exceptionally rich in coal and copper-gold resources - with world-class deposits like Tavan Tolgoi, Oyu Tolgoi and Erdenet. Thus, the mining industry has a crucial importance for the national economy, yet most of the country remain very underexplored. Within today's global tectonics, an acceptable understanding of metal enrichments - including leaching, the internal hydrostatic-hydraulic pumping system, and surface emplacement mechanisms - has remained unresolved. However, a broader view of the structural situation in the Mongolia-China region shows a close link between orientation of elongate sedimentary basins, important mineral belts, and the fundamental orthogonal fracture/fault system. In the east the tectonic trend is dominantly northeast, while it is northwest in western areas. The main east Mongolian graphite deposits have northeast structural trends like numerous regional Cu and Au belts. A new theory of the earth, Global Wrench Tectonics, offers an exciting approach to better understanding the various facets of Earth's geological history and its surface resources. Earth’s degassing, dynamo-tectonic consequences, inertia-driven crustal wrench tectonics, as well as surface products such as water, hydrocarbons and ore deposits are given a coherent system explanation. Many hydrocarbons are products from the interior of our slowly degassing Earth, with massive hydrocarbon fields such as Songliao and the Yamal megaproject producing from the basement. Crustal thinning in the Songliao region is about the same as in southeast Mongolia, suggesting that they may have had similar degassing and crustal evolution histories. As such, it is not unlikely that the underexplored Mesozoic basins of southeast Mongolia - particularly at the deepest levels and/or in the adjacent crystalline basement - may have important hydrocarbon potential.

Online Detection of Component Concentration in Synthetic Sodium Aluminate Solution Using Orthogonal Regression and BP Neural Network

Online Detection of Component Concentration in Synthetic Sodium Aluminate Solution Using Orthogonal Regression and BP Neural Network

Study on the Mechanism and Regulation Method of Longitudinal Penetration of Hydraulic Fractures in Multilayered Shale

Summary Shale reservoirs have longitudinally developed multilayered weak surfaces. The strong geological discontinuity and the stress heterogeneity caused by it lead to the complicated morphology of hydraulic fracture propagation, and the longitudinal propagation mechanism of the hydraulic fracture is still unclear. The extended finite element 3D numerical model of the single-cluster fracture and multicluster fracture extension has been established. The effects of vertical stress difference, bonding strength of bedding plane, fracturing fluid displacement, fracturing fluid viscosity, and cluster spacing on fracture propagation morphology are analyzed by numerical examples. The results show that as the vertical stress difference and the bonding strength of the bedding plane increase, the bedding plane becomes more difficult to activate, and the fractures are more likely to realize the longitudinal penetration. As the cluster spacing decreases, the interfracture interference becomes stronger, and the hydraulic fractures are more likely to activate the bedding plane and form the orthogonal network fracture. At a high injection rate, the fracture passes easily through the layer and activates the bedding plane. Low-viscosity fracturing fluid is conducive to the activation of the bedding plane, and high-viscosity fracturing fluid can better achieve fracture penetration. Based on the research results, the fracturing parameters of Well X-1 are optimized, and the fracture monitoring results are in good agreement with the design objectives. This study reveals the longitudinal penetration mechanism of multilayered shale hydraulic fractures and provides a reference for the optimization of hydraulic fracturing parameters of multilayered shale.

TSOANet: Time-Sensitive Orthogonal Attention Network for medical event prediction

Medical Event Prediction (MEP) based on Electronic Medical Records (EMR) is an essential and valuable task for healthcare. For a patient, information in the EMR can be organized into a structured sequence, consisting of multiple visits each with details about visit time and various types of medical events. As the time intervals between neighboring visits are irregular and the medical events at different visits can vary significantly, MEP based on EMR is still challenging. Many studies have been proposed to model the irregular time intervals, relations among different types of medical events within each visit and relations among medical events across visits, and reported exciting results. However, most of these studies focus on two out of the three aspects mentioned above, with only a few addressing all the three aspects simultaneously. In this study, we propose a novel network, the Time-Sensitive Orthogonal Attention Network (TSOANet), which can fully utilize the irregular time intervals, relations among different types of medical events within and across visits. In particular, we design two key components: (1) Time-Sensitive Block, used to model the time intervals at both local and global levels to determine the impact of each visit in EMR; (2) Orthogonal Attention Block, used to model relations among different types of medical events within each visit and across visits in two axes, that is, event axis and time axis. Extensive experiments on two public real-world EMR datasets demonstrate that TSOANet outperforms the state-of-the-art models for various prediction tasks, thereby verifying the effectiveness of our approach. The source code of TSOANet is released at https://github.com/chh13502/TSOANet.

Chaotic discrete Fresnel transform for secure OFDM-PONs

A chaotic discrete Fresnel transform (DFnT) precoding scheme is proposed for physical-layer security enhancement in orthogonal frequency division multiplexing passive optical networks (OFDM-PONs). A multi-fold OFDM data encryption is applied in terms of the chaotic scale factor, and chaotic row/column permutations in the DFnT matrix, and the chaos sensitivity creates a huge key space of 10138. The peak-to-average power ratio (PAPR) performances of OFDM signals associated with a scale factor in DFnT are fully analyzed. A 38.2 Gb/s OFDM signal transmission with chaotic DFnT is successfully demonstrated over a 20 km standard single-mode fiber (SSMF), where the gain of receiver sensitivity is ∼2 dB (BER @10−3), compared with OFDM without precoding.

Fast reconstruction of boiler numerical physical field based on proper orthogonal decomposition and conditional deep convolutional generative adversarial networks

ABSTRACT Obtaining real-time physical field data inside the boiler furnace is crucial for combustion diagnostics and optimization. In this work, we propose a fast reconstruction technique for the physical fields in a tangentially-fired coal boiler. The method combines the Proper Orthogonal Decomposition (POD) and Conditional Deep Convolutional Generative Adversarial Networks (cDCGAN) based on the boiler physical field obtained by computational fluid dynamics (CFD). Firstly, CFD simulations are performed using data from the boiler Distributed Control System (DCS). After verifying the CFD results, we expanded the simulation to 105 operational condition cases sample sets. Then, the physical fields of the sample set undergo POD to determine the appropriate number of POD bases, thereby ensuring the accuracy of the reconstruction. Finally, the cDCGAN method is employed to establish the correlation between the POD bases and the operational conditions of the boiler. Results show that the POD-cDCGAN method achieves rapid and accurate reconstruction of temperature and velocity fields, with mean relative error (MRE) below 1.5% and root relative squared error (RRSE) below 5.7%. The reconstruction time is reduced to 10.57s, significantly faster than direct CFD methods. The POD-cDCGAN method offers an efficient approach for online operational diagnosis and digital twin construction of boiler equipment.

Spirobifluorene‐Based Porous Organic Salts: Their Porous Network Diversification and Construction of Chiral Helical Luminescent Structures

AbstractPorous organic salts (POSs) are organic porous materials assembled via charge‐assisted hydrogen bonds between strong acids and bases such as sulfonic acids and amines. To diversify the network topology of POSs and extend its functions, this study focused on using 4,4′,4′′,4′′′‐(9,9′‐spirobi[fluorene]‐2,2′,7,7′‐tetrayl)tetrabenzenesulfonic acid (spiroBPS), which is a tetrasulfonic acid comprising a square planar skeleton. The POS consisting of spiroBPS and triphenylmethylamine (TPMA) (spiroBPS/TPMA) was constructed from the two‐fold interpenetration of an orthogonal network with pts topology, which has not been reported in conventional POSs, owing to the shape of the spirobifluorene backbone. Furthermore, combining tris(4‐chlorophenyl)methylamine (TPMA‐Cl) and tris(4‐bromophenyl)methylamine (TPMA‐Br), which are bulkier than TPMA owing to the introduction of halogens at the p‐position of the phenyl groups with spiroBPS allows us to construct novel POSs (spiroBPS/TPMA‐Cl and spiroBPS/TPMA‐Br). These POSs were constructed from a chiral helical network with pth topology, which was induced by the steric hindrance between the halogens and the curved fluorene skeleton. Moreover, spiroBPS/TPMA‐Cl with pth topology exhibited circularly polarized luminescence (CPL) in the solid state, which has not been reported in hydrogen‐bonded organic frameworks (HOFs).

Spirobifluorene-Based Porous Organic Salts: Their Porous Network Diversification and Construction of Chiral Helical Luminescent Structures.

Porous organic salts (POSs) are organic porous materials assembled via charge-assisted hydrogen bonds between strong acids and bases such as sulfonic acids and amines. To diversify the network topology of POSs and extend its functions, this study focused on using 4,4',4'',4'''-(9,9'-spirobi[fluorene]-2,2',7,7'-tetrayl)tetrabenzenesulfonic acid (spiroBPS), which is a tetrasulfonic acid comprising a square planar skeleton. The POS consisting of spiroBPS and triphenylmethylamine (TPMA) (spiroBPS/TPMA) was constructed from the two-fold interpenetration of an orthogonal network with pts topology, which has not been reported in conventional POSs, owing to the shape of the spirobifluorene backbone. Furthermore, combining tris(4-chlorophenyl)methylamine (TPMA-Cl) and tris(4-bromophenyl)methylamine (TPMA-Br), which are bulkier than TPMA owing to the introduction of halogens at the p-position of the phenyl groups with spiroBPS allows us to construct novel POSs (spiroBPS/TPMA-Cl and spiroBPS/TPMA-Br). These POSs were constructed from a chiral helical network with pth topology, which was induced by the steric hindrance between the halogens and the curved fluorene skeleton. Moreover, spiroBPS/TPMA-Cl with pth topology exhibited circularly polarized luminescence (CPL) in the solid state, which has not been reported in hydrogen-bonded organic frameworks (HOFs).

Orthogonal Neural Network: An Analytical Model for Deep Learning

In the current deep learning model, the computation between each feature and parameter is defined in the real number field. This, together with the nonlinearity of the deep learning model, makes it difficult to analyze the relationship between the values of the computational process and the original features from computation in the real number field. We extend the operational rules of the deep learning model in space and propose the orthogonal neural network (ONN) model, in which the features are set orthogonally to each other in space by “modulating” each input feature of the deep learning model to different orthogonal bases. Because the modulated numerical features are orthogonal to each other, they can be separated from the computations of the ONN model. By “demodulating” the model during and after the calculation, we can obtain a numerical relationship between the results and the original features, which can further provide theoretical and computational support for our analysis of the model. Finally, we compute the weights for each input feature as an interpretable deep learning approach, and describe how the model focuses attention on each feature based on the application of the orthogonal neural network model on two typical models: convolutional neural networks and graph neural networks.

Orthogonal Capsule Networks With Positional Information Preservation and Lightweight Feature Learning.

Both transformer and convolutional neural network (CNN) models require supplementary elements to acquire positional information. To address this issue, we propose a novel orthogonal capsule network (OthogonalCaps) that preserves location information during lightweight feature learning. The proposed network simplifies complex training processes and enables end-to-end training for object detection tasks. Specifically, there is no need to solve the regression problem of positions and the classification problem of objects separately, nor is there a need to encode the positional information as an additional token, as in transformer models. We generate the next capsule layer via orthogonality-based dynamic routing, which reduces the number of parameters and preserves positional information via its voting mechanism. Moreover, we propose Capsule ReLU as an activation function to avoid the problem of gradient vanishing and to facilitate capsule normalization across various scales, thus empowering OrthogonalCaps to better adapt to objects of diverse scales. The orthogonal capsule network (CapsNet) demonstrates an accuracy and run-time performance on a par with those of Faster R-CNN on the VOC dataset. Our network outperforms the baseline approach in detecting small-scale samples. The simulation results suggest that the proposed network surpasses other capsule network models in achieving a favorable balance between parameters and accuracy. Furthermore, an ablation experiment indicates that both Capsule ReLU and orthogonality-based dynamic routing play essential roles in enhancing the classification performance. The training code and pretrained models are available at https://github.com/l1ack/OrthogonalCaps.

Multi-view 3D human pose estimation based on multi-scale feature by orthogonal projection

Aiming at the problems of inaccurate estimation results, complicated matching of feature information in different views and poor robustness of the network model in complex scenes, a multi-view multi-person 3D human pose estimation model with multi-scale feature orthogonal projection is proposed, which includes a multi-scale orthogonal projection fusion network and an orthogonal feature ascending dimension network. Firstly, the multi-scale orthogonal projection fusion network performs orthogonal projection of features at multiple scales, using the residual structure to fuse features in the same plane separately, simplifying the feature learning difficulty and reducing the feature loss due to projection. Then, it is fed into the orthogonal feature ascending dimension network to reconstruct higher level 3D features using trilinear interpolation and deconvolution to improve the expressiveness of the model, and finally fed to the backbone network to supplement the information of the high-dimensional features, and the network regresses according to the different stages of the task to obtain the 3D human pose. The experimental results show that the Percentage of 3D Correct Parts is improved on the Campus and Shelf datasets, and the Mean Per Joint Position Error is reduced on the CMU Panoptic dataset and the average accuracy is improved at a smaller threshold compared to the previous method. The prediction results are also better than the previous method by reducing the perspective input on the trained model. The proposed method not only effectively estimates the 3D human pose, but also improves the prediction accuracy and enhances the robustness of the network model.

ORNet: Orthogonal Re-Parameterized Networks for Fast Pedestrian and Vehicle Detection

ORNet: Orthogonal Re-Parameterized Networks for Fast Pedestrian and Vehicle Detection