Hierarchical Interactions Research Articles

Uncertainty Quantification and Interpretability for Clinical Trial Approval Prediction.

Background: Clinical trial is a crucial step in the development of a new therapy (e.g., medication) and is remarkably expensive and time-consuming. Forecasting the approval of clinical trials accurately would enable us to circumvent trials destined to fail, thereby allowing us to allocate more resources to therapies with better chances. However, existing approval prediction algorithms did not quantify the uncertainty and provide interpretability, limiting their usage in real-world clinical trial management. Methods: This paper quantifies uncertainty and improves interpretability in clinical trial approval predictions. We devised a selective classification approach and integrated it with the Hierarchical Interaction Network, the state-of-the-art clinical trial prediction model. Selective classification, encompassing a spectrum of methods for uncertainty quantification, empowers the model to withhold decision-making in the face of samples marked by ambiguity or low confidence. This approach not only amplifies the accuracy of predictions for the instances it chooses to classify but also notably enhances the model's interpretability. Results: Comprehensive experiments demonstrate that incorporating uncertainty markedly enhances the model's performance. Specifically, the proposed method achieved 32.37%, 21.43%, and 13.27% relative improvement on area under the precision-recall curve over the base model (Hierarchical Interaction Network) in phase I, II, and III trial approval predictions, respectively. For phase III trials, our method reaches 0.9022 area under the precision-recall curve scores. In addition, we show a case study of interpretability that helps domain experts to understand model's outcome. The code is publicly available at https://github.com/Vincent-1125/Uncertainty-Quantification-on-Clinical-Trial-Outcome-Prediction. Conclusion: Our approach not only measures model uncertainty but also greatly improves interpretability and performance for clinical trial approval prediction.

Active Interdiction Defence Scheme Against False Data-Injection Attacks: A Stackelberg Game Perspective.

The security issue is of the fundamental importance for cyber-physical systems (CPSs) due to the vulnerability to cyber attacks. In this article, an active interdiction defence scheme is proposed as an extra safeguard against the strategic false data-injection (FDI) attacker. To tackle the coupling between the decision-making processes of the FDI attacker and the active interdiction defence scheme, a framework of dynamic Stackelberg game is formulated to characterize the hierarchical interaction due to their asymmetric information structures. Subsequently, the Stackelberg strategies of both players are constructed. Furthermore, a sufficient condition is derived to guarantee asymptotic stability of the closed-loop system with the FDI attack and the active interdiction defence scheme. Finally, two simulation examples are provided to illustrate the correctness and effectiveness of the proposed active interdiction defence scheme.

DHI-Net: A Novel Detail-Preserving and Hierarchical Interaction Network For Building Extraction

DHI-Net: A Novel Detail-Preserving and Hierarchical Interaction Network For Building Extraction

Elastomer Composites with High Damping and Low Thermal Resistance via Hierarchical Interactions and Regulating Filler.

Modern microelectronics and emerging technologies such as wearable electronics and soft robotics require elastomers to integrate high damping with low thermal resistance to avoid damage caused by vibrations and heat accumulation. However, the strong coupling between storage modulus and loss factor makes it generally challenging to simultaneously increase both thermal conductance and damping. Here, a strategy of introducing hierarchical interaction and regulating fillers in polybutadiene/spherical aluminum elastomer composites is reported to simultaneously achieve extraordinary damping ability of tan δ>1.0 and low thermal resistance of 0.15cm2 KW-1 , which surpasses state-of-the-art elastomers and their composites. The enhanced damping is attributed to increased energy dissipation via introducing the hierarchical hydrogen bond interactions in polybutadiene networks and the addition of spherical aluminum, which also functions as a thermally conductive filler to achieve low thermal resistance. As a proof of concept, the polybutadiene/spherical aluminum elastomer composites are used as thermal interface materials, showing effective heat dissipation for electronic devices in vibration scenarios. The combination of outstanding damping performance and extraordinary heat dissipation ability of the elastomer composites may create new opportunities for their applications in electronics.

When Nash Meets Stackelberg

This article introduces a class of Nash games among Stackelberg players (NASPs), namely, a class of simultaneous noncooperative games where the players solve sequential Stackelberg games. Specifically, each player solves a Stackelberg game where a leader optimizes a (parametrized) linear objective function subject to linear constraints, whereas its followers solve convex quadratic problems subject to the standard optimistic assumption. Although we prove that deciding if a NASP instance admits a Nash equilibrium is generally a [Formula: see text]-hard decision problem, we devise two exact and computationally efficient algorithms to compute and select Nash equilibria or certify that no equilibrium exists. We use NASPs to model the hierarchical interactions of international energy markets where climate change aware regulators oversee the operations of profit-driven energy producers. By combining real-world data with our models, we find that Nash equilibria provide informative, and often counterintuitive, managerial insights for market regulators. This paper was accepted by Chung Piaw Teo, optimization. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2022.03418 .

Random hierarchical model for estimation of wheat yield in the North China Plain at different spatial scales

Accurate and large-scale wheat yield prediction in the North China Plain (NCP) can provide necessary information for agricultural policies and agricultural trade. Many studies have presented wheat yield estimation methods by using existing machine learning methods and remote sensing (RS) or environmental data. However, these methods only blindly input multi-source data into the model and do not consider the hierarchical relationships and interaction between different data types. In addition, there has been scant attention paid to the consistency of yield estimation models across varying spatial scales. To address these problems, a novel dynamic yield estimation model, known as random hierarchical model (RHM), which takes into account the hierarchical relationship of multi-source data, is constructed to estimate the wheat yield in the NCP. First, the time interval of wheat growth period is refined by using the 24 solar terms in China, and a time series multi-source dataset of climate, soil, and RS is constructed. Second, the hierarchical linear model is used to layer multi-source data and randomly select environmental and RS features with multiple time intervals. Multiple hierarchical models are constructed and optimized for integration, and the interrelationships between data collected at different levels are fully utilized, which can improve the accuracy of the yield estimation models in interannual and large-scale applications. Finally, the RHM at different spatial scales is cross-verified by using the measured and statistical data of the NCP for 4 years. The results indicated that the error of the RHM estimation is smaller than that of widely used machine learning models at different spatial scales of field-level measurement data (R2 = 0.52, nRMSE = 16.43%), county-level measurement integration data (R2 = 0.62, nRMSE = 12.83%), and county-level official statistics (R2 = 0.68, nRMSE = 11.41%). Our proposed RHM that considers the hierarchical structure of multi-source data is a reliable and a promising method for improving yield estimation. In addition, heterogeneity in the hierarchical relationships is observed between the different types of data in the RHM at different spatial scales, resulting in differences in the optimal lead time for estimating yield and the importance of key driving factors in the model, indicating that the cross-spatial scale applications of the model are not allowed. This study provides insights for large-scale wheat yield estimation and yield response to different environments and provided evidence and explanation for the prohibition of generalization of models at different spatial scales.

Protists regulate microbially mediated organic carbon turnover in soil aggregates

AbstractSoil protists, the major predator of bacteria and fungi, shape the taxonomic and functional structure of soil microbiome via trophic regulation. However, how trophic interactions between protists and their prey influence microbially mediated soil organic carbon turnover remains largely unknown. Here, we investigated the protistan communities and microbial trophic interactions across different aggregates‐size fractions in agricultural soil with long‐term fertilization regimes. Our results showed that aggregate sizes significantly influenced the protistan community and microbial hierarchical interactions. Bacterivores were the predominant protistan functional group and were more abundant in macroaggregates and silt + clay than in microaggregates, while omnivores showed an opposite distribution pattern. Furthermore, partial least square path modeling revealed positive impacts of omnivores on the C‐decomposition genes and soil organic matter (SOM) contents, while bacterivores displayed negative impacts. Microbial trophic interactions were intensive in macroaggregates and silt + clay but were restricted in microaggregates, as indicated by the intensity of protistan‐bacterial associations and network complexity and connectivity. Cercozoan taxa were consistently identified as the keystone species in SOM degradation‐related ecological clusters in macroaggregates and silt + clay, indicating the critical roles of protists in SOM degradation by regulating bacterial and fungal taxa. Chemical fertilization had a positive effect on soil C sequestration through suppressing SOM degradation‐related ecological clusters in macroaggregate and silt + clay. Conversely, the associations between the trophic interactions and SOM contents were decoupled in microaggregates, suggesting limited microbial contributions to SOM turnovers. Our study demonstrates the importance of protists‐driven trophic interactions on soil C cycling in agricultural ecosystems.

Ultrafast Self-Healing, Superstretchable, and Ultra-Strong Polymer Cluster-Based Adhesive Based on Aromatic Acid Cross-Linkers for Excellent Hydrogel Strain Sensors.

Synthetic hydrogel strain sensors rarely exhibit a comprehensive combination of mechanical properties such as ultra-stretchability, ultrafast self-healing, and high sensitivity. Herein, seven small molecule enhanced mechanical behaviors of polymer-cluster based hydrogels are demonstrated. The oxidized polyethyleneimine/polymeric acrylic acid (ohPEI/PAA) hydrogels with aromatic formic acids as supramolecular cross-linkers are prepared by simultaneous formation of ohPEI polymer clusters and PAA upon the addition of ammonium persulfate. The optimized hydrogel adhesive exhibits comprehensive excellent properties, such as high extensibility (up to 12298%), real-time mechanical self-healing capability (<1s, 93% efficiency), high uniformity, underwater adhesivity, and water-sealing ability. The proper binding strength of hydrogel and skin (47kPa) allows the hydrogel to be utilized as highly sensitive (gauge factor:16.08), highly conductive (2.58mScm-1), and underwater strain sensors. Specially, the adhesive strength of the adhesive to wood after dehydration is extremely high, reaching up to 29.59MPa. Additionally, when glycerol is introduced, the obtained gel maintains the physical properties even at harsh-temperature conditions (-40to 80 °C). It presents that multiple and hierarchical non-covalent interactions including multiple hydrogen bonding interactions, π-π stacking, electrostatic interactions, and dipole-dipole interactions of polymer clusters, allow for the energy dissipation and contribute to the excellent performance of the hydrogel.

MMFIL-Net: Multi-level and multi-source feature interactive lightweight network for polyp segmentation

Accuracy, generalized performance, and model size are critical considerations for designing a real-time polyp segmentation model. However, existing techniques primarily focus on accuracy and do not consider the other two vital characteristics. This paper proposes MMFIL-Net, a novel polyp segmentation technique. As part of MMFIL-Net, the Hierarchical Multi-source Feature Interaction Module (HMFIM) comprises Multi-source Feature Interaction Blocks (MFIB). MFIB manipulates multi-level and multi-sourced features to reduce the gap between low and high-level feature maps to achieve generalized performance. Additionally, the Multiple Receptive Field Feature Interaction Block (MRFFIB) targets the issues of segmenting polyps of different sizes. Finally, Dual Source Attention Fusion Block (DSAFB) is introduced to deal with hazy boundary information for early-stage polyps detection and segmentation. The proposed model outperformed existing lightweight models in conducted evaluation on different datasets. In addition to the achieved generalized performance and higher accuracy, the proposed model presents a significant reduction in the model size than existing approaches. The proposed model only contains 6.68 million parameters and has 4.32G MACs (Multiply–Accumulate Operations), which is better than the current approaches.

Towards sturdy and sensable pressure-sensitive adhesive through hierarchical supramolecular interaction

Endowing a robust pressure-sensitive adhesive (PSA) with sensable properties is of great significance for in-situ stress detection and information encryption in the fields of electronics, energy storage, flexible sensing, etc. However, it remains great challenge due to the difficulty in balancing interfacial wetting and cohesive strength. Herein, a microphase-separated strategy is proposed to construct an ionogel with a lower modulus of 1.96MPa, a strength of 728kPa as well as a remarkable toughness of 2258.9kJ/m3, which can be used as a sturdy PSA bonded to various substrates (metals, polar plastics, non-polar plastics) under gentle pressure. The comparable modulus and cohesive strength give it an excellent adhesion strength of 1340kPa, which far exceeds most of reported high-performance PSAs. Furthermore, due to the orientation of a large number of ionic groups, the adhesion strength increases by 31.3% once a voltage of 20V is applied. Finally, the sensitive force-resistance response of such PSA that can be used for encrypted messaging was demonstrated.

Hierarchically ordered coherent interfaces-driven ultrahigh specific-strength and toughness in a nano-martensite titanium alloy

Precipitates in alloys are traditionally considered as dislocation obstacles that often lead to high stress concentration and even microcracks owing to the strain incompatibility of semi-coherent interfaces, a cause of progressive strain localization and the origin of the strength-toughness conflict in engineering materials, such as titanium (Ti) alloys. Based on the metastability engineering, here we architect hierarchically ordered coherent interfaces for strength-toughness optimization through densely dispersed nanomartensites in a ductile Ti-Cr-Zr-Al alloy with ultrahigh specific-strength and superior fracture toughness. It is unveiled that these ordered coherent interfaces simultaneously serve as dislocation obstacles and sources, leading to a sustainable and self-hardening deformation mechanism via hierarchical nanomartensite-dislocation interactions for ultra-high strength and toughness of Ti alloys. These nanomartensites are thermally stable at elevated temperature less than 400 °C, above which tempering-induced ductile-to-brittle transition occurs due to the decomposition of hierarchically ordered nanomartensites and the spheroidization of prior β lamellae. The design strategy of hierarchically ordered coherent interfaces confers our cost-effective nanomartensite Ti alloys an unprecedented combination of strength, ductility and toughness, which provides a new pathway in the microstructural design for strong and ductile structural materials with superior fracture resistance.

Dual-graph hierarchical interaction network for referring image segmentation

Referring Image Segmentation (RIS) aims to extract the object or stuff from an image according to the given natural language expression. As a representative multi-modal reasoning task, the main challenge of RIS lies in accurately understanding and aligning two types of heterogeneous data (i.e. image and text). Existing methods typically conduct this task via inexplicit cross-modal fusion toward the visual and linguistic features, which are separately extracted from different encoders and hard to capture accurate image–text alignment information due to their distinct latent representation structures. In this paper, we propose a Dual-Graph Hierarchical Interaction Network (DGHIN) to facilitate the explicit and comprehensive alignment between the image and text data. Firstly, two graphs are separately built for the initial visual and linguistic features extracted with different pre-trained encoders. By means of graph reasoning, we obtain a unified representation structure for different modalities to capture the intra-modal entities and their contexts, where each projected node incorporates the long-range dependencies into the latent representation. Then, the Hierarchical Interaction Module (HIM) is applied to the visual and linguistic graphs to extract comprehensive inter-modal correlations from the entity level and graph level, which not only capture the corresponding keywords and visual patches but also draws the whole sentence closer to the image region with the consistent context in the latent space. Extensive experiments on RefCOCO, RefCOCO+, G-Ref, and ReferIt demonstrate that the proposed DGHIN outperforms many state-of-the-art methods. Code is available at https://github.com/ZhaofengSHI/referring-DGHIN.

Weakly guided attention model with hierarchical interaction for brain CT report generation

Brain Computed Tomography (CT) report generation, which aims to assist radiologists in diagnosing cerebrovascular diseases efficiently, is challenging in feature representation for dozens of images and language descriptions with several sentences. Existing report generation methods have achieved significant achievement based on the encoder–decoder framework and attention mechanism. However, current research has limitations in solving the many-to-many alignment between the multi-images of Brain CT imaging and the multi-sentences of Brain CT report, and fails to attend to critical images and lesion areas, resulting in inaccurate descriptions. In this paper, we propose a novel Weakly Guided Attention Model with Hierarchical Interaction, named WGAM-HI, to improve Brain CT report generation. Specifically, WGAM-HI conducts many-to-many matching for multiple visual images and semantic sentences via a hierarchical interaction framework with a two-layer attention model and a two-layer report generator. In addition, two weakly guided mechanisms are proposed to facilitate the attention model to focus more on important images and lesion areas under the guidance of pathological events and Gradient-weighted Class Activation Mapping (Grad-CAM) respectively. The pathological event acts as a bridge between the essential serial images and the corresponding sentence, and the Grad-CAM bridges the lesion areas and pathology words. Therefore, under the hierarchical interaction with the weakly guided attention model, the report generator generates more accurate words and sentences. Experiments on the Brain CT dataset demonstrate the effectiveness of WGAM-HI in attending to important images and lesion areas gradually, and generating more accurate reports.

Healable and recyclable multiblock polyurethanes with mechanical performance Tailorability based on hierarchical phase separation and dynamic bond interaction

Thermoplastic polyurethanes (TPUs) have been demonstrated high versatility in widespread applications. In view of today's sustainable society, the advanced TPUs which combine both excellent mechanical properties and functionalities such as the self-healing and recyclability can extend utilization scope and service life. Herein, we reported a multiblock polyurethane with outstanding mechanical performance, self-healing capability and recyclability based on strong microphase-separated biocompatible polycaprolactone (PCL) semicrystalline segments and ultra-highly flexible polydimethylsiloxane (PDMS) segments. The mechanical performance of the designed PUs can be tuned from resilient elastomer to shape memory plastics by varying the molecular parameter of segments, and the fast self-healing was achieved through the synergetic effects of dynamic disulfide bond and hydrogen bonding interaction. The PUs with high PCL contents exhibit high tensile strength and high toughness (high up to 45 MPa, 511 MJ m−3, respectively) after thermal treatment and show shape memory effects, while these with high PDMS contents perform as ligament elastomer. We demonstrated the principles of high mechanical performance in the phase-separated multiblock PU, in which the hierarchical nanostructures including the microphase separation between PDMS and PCL blocks, the nanophase segregation of the carbamate segments as well as the semi-crystalline behavior of PCL segments endow these healable and recyclable PUs with high strength, high stretchability and high toughness. This work demonstrated tremendous potentials of utilizing strong microphase separated segments and dynamic bond chain extenders to design functional thermoplastic PU materials.

RAGraph: A Region-Aware Framework for Geo-Distributed Graph Processing

In many global businesses of multinational enterprises, graph-structure data is usually geographically distributed in different regions to support low-latency services. Geo-distributed graph processing suffers from the Wide Area Networks (WANs) with scarce and heterogeneous bandwidth, thus essentially differs from traditional distributed graph processing. In this paper, we propose RAGraph, a Region-Aware framework for geo-distributed graph processing. At the core of RAGraph, we design a region-aware graph processing framework that allows advancing inefficient global updates locally and enables sensible coordination-free message interactions. RAGraph also contains an adaptive hierarchical message interaction engine to switch interaction modes adaptively based on network heterogeneity and fluctuation, and a discrepancy-aware message filtering strategy to filter important messages. Finally, the experiments show that RAGraph can achieve 1.69X - 40.53X speedup and 20.9% - 97% WAN cost reduction compared with state-of-the-art systems.

A network coupling approach to detecting hierarchical linkages between science and technology

AbstractDetecting science–technology hierarchical linkages is beneficial for understanding deep interactions between science and technology (S&amp;T). Previous studies have mainly focused on linear linkages between S&amp;T but ignored their structural linkages. In this paper, we propose a network coupling approach to inspect hierarchical interactions of S&amp;T by integrating their knowledge linkages and structural linkages. S&amp;T knowledge networks are first enhanced with bidirectional encoder representation from transformers (BERT) knowledge alignment, and then their hierarchical structures are identified based on K‐core decomposition. Hierarchical coupling preferences and strengths of the S&amp;T networks over time are further calculated based on similarities of coupling nodes' degree distribution and similarities of coupling edges' weight distribution. Extensive experimental results indicate that our approach is feasible and robust in identifying the coupling hierarchy with superior performance compared to other isomorphism and dissimilarity algorithms. Our research extends the mindset of S&amp;T linkage measurement by identifying patterns and paths of the interaction of S&amp;T hierarchical knowledge.

Hierarchical visual-semantic interaction for scene text recognition

Proper interaction between visual and semantic features is crucial to obtain a powerful feature representation for scene text recognition (STR). The existing interaction methods usually treat visual and semantic features as distinct tokens and use transformers to learn contextual information and prior language knowledge, and they achieve promising performance for STR task. However, there still remain several issues needed to be further addressed such as the imbalance number and mis-alignment between visual and semantic features, and the necessarily of stacking transformers to a progressive improvement in accuracy. To this aim, this paper proposes a novel interaction manner namely hierarchical visual-semantic interaction (HVSI) which contains three novel modules including a hierarchical visual-semantic interaction module, fusion module, and visual-semantic alignment module. The hierarchical visual-semantic interaction module employs multiple visual-semantic interaction blocks in various scales to enhance the representation power of visual features and semantic features. To better exploit multi-scale visual and semantic features, the fusion module is introduced to fuse multiple semantic features based on attention mechanisms. Furthermore, our HVSI presents a simple plug-in block named visual-semantic alignment module to alleviate mis-alignment of semantic features by mapping them into a unified semantic space, which helps improve the performance of HVSI. Extensive experiments on multiple benchmarks including English and Chinese text recognition datasets show that our method obtains state-of-the-art or competitive performances.

Descending GABAergic pathway links brain sugar-sensing to peripheral nociceptive gating in Drosophila

Although painful stimuli elicit defensive responses including escape behavior for survival, starved animals often prioritize feeding over escape even in a noxious environment. This behavioral priority is typically mediated by suppression of noxious inputs through descending control in the brain, yet underlying molecular and cellular mechanisms are incompletely understood. Here we identify a cluster of GABAergic neurons in Drosophila larval brain, designated as SEZ-localized Descending GABAergic neurons (SDGs), that project descending axons onto the axon terminals of the peripheral nociceptive neurons and prevent presynaptic activity through GABAB receptors. Remarkably, glucose feeding to starved larvae causes sustained activation of SDGs through glucose-sensing neurons and subsequent insulin signaling in SDGs, which attenuates nociception and thereby suppresses escape behavior in response to multiple noxious stimuli. These findings illustrate a neural mechanism by which sugar sensing neurons in the brain engages descending GABAergic neurons in nociceptive gating to achieve hierarchical interaction between feeding and escape behavior.

Stretchable and hydrophobic eutectogel for underwater human health monitoring based on hierarchical dynamic interactions

Stretchable and hydrophobic eutectogel for underwater human health monitoring based on hierarchical dynamic interactions

Value of C-peptide-based insulin resistance index for evaluating correlation between insulin resistance and serum uric acid level in individuals undergoing health examination

To investigate the value of C-peptide-based insulin resistance index in evaluating the correlation between insulin resistance and serum uric acid (Ua) level in subjects undergoing health examination. The data of 46 017 subjects undergoing health examination were retrospectively collected from the Second Medical Center of PLA General Hospital from January, 2017 to December, 2021. The subjects were divided into Ua≤420 μmol/L group and Ua>420 μmol/L group for comparison of HOMA insulin resistance index (HOMA2-IR) and HOMA insulin resistance-C peptide (HOMA2 IR-CP). The correlations of HOMA2-IR and HOMA2 IR-CP with Ua level were analyzed using Pearson correlation analysis and linear regression analysis. Hierarchical interaction analysis was conducted to assess the differences in the association between insulin resistance index and Ua level in different subgroups. The ROC curve was used to evaluate the predictive ability of insulin resistance index for an increased Ua level. The levels of HOMA2-IR and HOMA2 IR-CP were significantly lower in Ua≤420 μmol/L group than in Ua>420 μmol/L group. Univariate Pearson correlation analysis showed a weak correlation of HOMA2-IR with Ua (r=0.262, P<0.001) and moderate correlation of HOMA2 IR-CP with Ua (r=0.409, P<0.001). Multivariate linear regression analysis, after adjustment for confounding factors, demonstrated that HOMA2-IR (R2=0.445, P<0.001) and HOMA2 IR-CP (R2=0.461, P<0.001) were both factors affecting Ua level. Hierarchical interaction analysis showed that the association of insulin resistance index with Ua level varied significantly with gender, age, and glucose metabolism (P<0.001). ROC curve showed that the areas under the curve predicted an increased Ua level by HOMA2-IR and HOMA2 IR-CP were 0.662 and 0.722, respectively. HOMA2 IR-CP is a more accurate indicator for assessing the correlation between insulin resistance and Ua level.