Downstream Network Research Articles

Pain in Parkinson's disease: a neuroanatomy-based approach.

Parkinson's disease is a progressive neurodegenerative disorder characterized by the deposition of misfolded alpha-synuclein in different regions of the central and peripheral nervous system. Motor impairment represents the signature clinical expression of Parkinson's disease. Nevertheless, non-motor symptoms are invariably present at different stages of the disease and constitute an important therapeutic challenge with a high impact for the patients' quality of life. Among non-motor symptoms, pain is frequently experienced by patients, being present in a range of 24-85% of Parkinson's disease population. Moreover, in more than 5% of patients, pain represents the first clinical manifestation, preceding by decades the exordium of motor symptoms. Pain implies a complex biopsychosocial experience with a downstream complex anatomical network involved in pain perception, modulation, and processing. Interestingly, all the anatomical areas involved in pain network can be affected by a-synuclein pathology, suggesting that pathophysiology of pain in Parkinson's disease encompasses a 'pain spectrum', involving different anatomical and neurochemical substrates. Here the various anatomical sites recruited in pain perception, modulation and processing are discussed, highlighting the consequences of their possible degeneration in course of Parkinson's disease. Starting from peripheral small fibres neuropathy and pathological alterations at the level of the posterior laminae of the spinal cord, we then describe the multifaceted role of noradrenaline and dopamine loss in driving dysregulated pain perception. Finally, we focus on the possible role of the intertwined circuits between amygdala, nucleus accumbens and habenula in determining the psycho-emotional, autonomic and cognitive experience of pain in Parkinson's disease. This narrative review provides the first anatomically driven comprehension of pain in Parkinson's disease, aiming at fostering new insights for personalized clinical diagnosis and therapeutic interventions.

Global solar photovoltaic industry network dynamics 2007–2023. Inter-organizational relationships as a source of competitive advantage?

Aiming a cleaner production in course of fighting the ongoing global warming, solar photovoltaic (PV) together with wind and hydro energy, indicate the most important industry segments in the transformation from fossils to renewable energy sources. During the last two decades, the solar PV industry experienced decisive changes of its global business network configurations where Chinese firms comparatively have gained competitive advantages. According to the business network theory, inter-organizational relationships indicate important vehicles for exchanging valuable resources such as advanced technology and privileged market knowledge aiming to gain leading industry positions.Based on a sample of globally leading solar PV manufacturers originated in Canada, China, Germany, South Korea, and the United States of America we conduct a detailed analysis and provide insights into solar PV industry upstream and downstream network dynamics examined for the period 2007–2023. We found that Chinese firms have focused more on upstream industry network relationships compared to their Western counterparts.Thus, we emphasize on the importance of building competitive regional solar PV industry networks instead of relying only on government subsidies which often come along with rather short-term effects.

Assessing efficiency and resilience in the Dewathang Thromde water distribution system

Dewathang Thromde, a subdivision of Samdrup Jongkhar Thromde, faces a significant challenge in its drinking water supply. The region relies on a gravitational water supply system from Lemtshorong and Zalakhe. This study aims to assess the current system and determine its adequacy for the future population. The study employs manual methods, such as bucket and stopwatch techniques, to determine discharge at the water source, complemented by field observations for pipe size measurements. Utilizing EPANET software and a Demand Driven Analysis (DDA) approach, distribution network’s pressure and velocity requirements are analyzed. Elevation data from the My Elevation mobile app, validated by Google Earth images, contribute to a comprehensive evaluation. During peak season, a discharge of 35.92 l/s is documented, contrasting with 9 l/s during the lean season. The population, recorded at 2644 in 2005, surged to 3881 in 2017, with a forecast anticipating 7170 people by 2035 using the Incremental Increase Method. Average water consumption per person is approximately 113.35 lpcd, exceeding the WHO standard, influenced by climatic conditions. With a population of approximately 4389 in 2021, the current water demand in Dewathang Thromde is 13.72 l/s, surpassing the discharge capacity of 9 l/s during the lean season. By 2035, the projected water demand is expected to reach 22.41 l/s, exceeding the current source discharge by over 100%. Simulation results for the upstream and downstream networks of the Break Pressure Tank (BPT) reveal pressure-related challenges, with over 20 percent of nodes experiencing excessive pressure along HDPE pipes. Bursting incidents and negative pressure at more than 10 percent of nodes emphasize the need for network improvements, especially during reduced source discharge. The final segment of the pipe network surpasses the prescribed maximum flow velocity due to the direct proportionality between velocity and pipe diameter. Smaller pipe sizes, while limiting velocity, paradoxically contribute to actual velocities exceeding specified limits. Addressing these challenges is crucial for ensuring a sustainable and reliable water supply system in Dewathang Thromde.

Based on BERT-wwm for Agricultural Named Entity Recognition

With the continuous advancement of information technology in the agricultural field, a large amount of unstructured agricultural textual information has been generated. This information is crucial for supporting the development of smart agriculture, making the application of named entity recognition in the agricultural field more urgent. In order to enhance the accuracy of agricultural entity recognition, this study utilizes the pre-trained BERT-wwm model for word embedding into the text. Additionally, a channel attention mechanism (CA) is introduced in the BILSTM-CRF downstream feature extraction network to comprehensively capture the contextual features of the text. Experimental results demonstrate that the proposed method significantly improves the performance of named entity recognition, with increased accuracy, recall, and F1 value. The successful implementation of this method provides reliable support for downstream tasks such as agricultural knowledge graph construction and question and answer systems and establishes a foundation for better understanding and utilization of agricultural textual information.

Effectiveness of Value-Added Input-Output Method in Upstream and Midstream Supply Chain Network of Sunflower Agro-industry

Abstract Sunflower (Helianthus annuus L) is an annual shrub that grows up to three meters high. Sunflower agro-industry has high potential economic value. Sunflower seeds produce oil that can be extracted and used as a raw material in the food, pharmaceutical, bioenergy, and cosmetic industries. The problem is the low availability of superior local seeds, so it must be imported. The impact of imports is the low value-added. Therefore, this study aims to determine and measure the value-added of sunflowers in the upstream and midstream networks. The methods are Hayami and BPS input-output, also aims to identify the effectiveness of these two methods.. The results of value-added on sunflower seeds, cooking oil, herbal oil, and animal feed by the Hayami and the BPS input-output each are IDR6,000 (40%) and IDR24,005,528 (20%), IDR4,900 (26%) and IDR41,377,013 (23%), IDR21,000 (60%) and IDR117,653,333 (38%), IDR3,500 (86%) and IDR23,996,000 (100%). it is better to calculate using the BPS method because it is more detailed and clear in all aspects of the sunflower agroindustry, while using the hayami method only focuses on sales and labor. In the future, this research will be developed by designing sustainability improvement to reduce environmental impacts; performing value-added calculations for derivative products in the downstream supply chain network; also developing a decision support system.

Arabidopsis transcription factor ANAC102 predominantly expresses a nuclear protein and acts as a negative regulator of methyl viologen-induced oxidative stress responses.

Plants, being sessile organisms, constantly need to respond to environmental stresses, often leading to the accumulation of reactive oxygen species (ROS). While ROS can be harmful, they also act as second messengers guiding plant growth and stress responses. Because chloroplasts are sensitive to environmental changes and are both a source and a target of ROS during stress conditions, they are important in conveying environmental changes to the nucleus, where acclimation responses are coordinated to maintain organellar and overall cellular homeostasis. ANAC102 has previously been established as a regulator of β-cyclocitral-mediated chloroplast-to-nucleus signaling, protecting plants against photooxidative stress. However, debates persist about where ANAC102 is located-in chloroplasts or in the nucleus. Our study, utilizing the genomic ANAC102 sequence driven by its native promoter, establishes ANAC102 primarily as a nuclear protein, lacking a complete N-terminal chloroplast-targeting peptide. Moreover, our research reveals the sensitivity of plants overexpressing ANAC102 to severe superoxide-induced chloroplast oxidative stress. Transcriptome analysis unraveled a dual role of ANAC102 in negatively and positively regulating genome-wide transcriptional responses to chloroplast oxidative stress. Through the integration of published data and our own study, we constructed a comprehensive transcriptional network, which suggests that ANAC102 exerts direct and indirect control over transcriptional responses through downstream transcription factor networks, providing deeper insights into the ANAC102-mediated regulatory landscape during oxidative stress.

Identification of ARF Genes and Elucidation of the Regulatory Effects of PsARF16a on the Dormancy of Tree Peony Plantlets.

The low survival rate of transplanted plantlets, which has limited the utility of tissue-culture-based methods for the rapid propagation of tree peonies, is due to plantlet dormancy after rooting. We previously determined that the auxin response factor PsARF may be a key regulator of tree peony dormancy. To clarify the mechanism mediating tree peony plantlet dormancy, PsARF genes were systematically identified and analyzed. Additionally, PsARF16a was transiently expressed in the leaves of tree peony plantlets to examine its regulatory effects on a downstream gene network. Nineteen PsARF genes were identified and divided into four classes. All PsARF genes encoded proteins with conserved B3 and ARF domains. The number of motifs, exons, and introns varied between PsARF genes in different classes. The overexpression of PsARF16a altered the expression of NCED, ZEP, PYL, GA2ox1, GID1, and other key genes in abscisic acid (ABA) and gibberellin (GA) signal transduction pathways, thereby promoting ABA synthesis and decreasing GA synthesis. Significant changes to the expression of some key genes contributing to starch and sugar metabolism (e.g., AMY2A, BAM3, BGLU, STP, and SUS2) may be associated with the gradual conversion of sugar into starch. This study provides important insights into PsARF functions in tree peonies.

MiR-146a alleviates inflammatory bowel disease in mice through systematic regulation of multiple genetic networks.

Inflammatory bowel disease (IBD) is a chronic disease involving multiple genes, and the current available targeted drugs for IBD only deliver moderate efficacy. Whether there is a single gene that systematically regulates IBD is not yet known. MiR-146a plays a pivotal role in repression of innate immunity, but its function in the intestinal inflammation is sort of controversy, and the genetic regulatory networks regulated by miR-146a in IBD has not been revealed. RT-qPCR was employed to detect the expression of miR-146a in IBD patients and in a mouse IBD model induced by dextran sulfate sodium (DSS), and then we generated a miR-146a knock-out mouse line with C57/Bl6N background. The disease activity index was scored in DSS-treated miR-146a deficiency mice and their wild type (WT) littermates. Bulk RNA-sequencing, RT-qPCR and immunostaining were done to illustrate the downstream genetic regulatory networks of miR-146a in flamed colon. Finally, the modified miR-146a mimics were used to treat DSS-induced IBD in miR-146a knock-out and WT IBD mice. We showed that the expression of miR-146a in the colon was elevated in dextran sulfate sodium (DSS)-induced IBD mice and patients with IBD. DSS induced dramatic body weight loss and more significant rectal bleeding, shorter colon length, and colitis in miR-146a knock-out mice than WT mice. The miR-146a mimics alleviated DSS-induced symptoms in both miR-146a-/- and WT mice. Further RNA sequencing illustrated that the deficiency of miR-146a de-repressed majority of DSS-induced IBD-related genes that cover multiple genetic regulatory networks in IBD, and supplementation with miR-146a mimics inhibited the expression of many IBD-related genes. Quantitative RT-PCR or immunostaining confirmed that Ccl3, Saa3, Csf3, Lcn2, Serpine1, Serpine2, MMP3, MMP8, MMP10, IL1A, IL1B, IL6, CXCL2, CXCL3, S100A8, S100A9, TRAF6, P65, p-P65, and IRAK1 were regulated by miR-146a in DSS induced IBD. Among them, MMP3, MMP10, IL6, IL1B, S100A8, S100A9, SERPINE1, CSF3, and IL1A were involved in the active stage of IBD in humans. Our date demonstrated that miR-146a acts as a top regulator in C57/BL6N mice to systematically repress multiple genetic regulatory networks involved in immune response of intestine to environment factors, and combinatory treatment with miR-146a-5p and miR-146a-3p mimics attenuates DSS-induced IBD in mice through down-regulating multiple genetic regulatory networks which were increased in colon tissue from IBD patients. Our findings suggests that miR-146a is a top inhibitor of IBD, and that miR-146a-5p and miR-146a-3p mimics might be potential drug for IBD.

Induction of let-7d-5p miRNA modulates aortic smooth muscle inflammatory signaling and phenotypic switching

Background and aimsActivation of vascular smooth muscle cell inflammation is recognised as an important early driver of vascular disease. We have previously identified the let-7 miRNA family as important regulators of inflammation in in vitro and in vivo models of atherosclerosis. Here we investigated a dual statin/let-7d-5p miRNA combination therapy approach to target human aortic SMC (HAoSMC) activation and inflammation. MethodsIn vitro studies using primary HAoSMCs were performed to investigate the effects of let-7d-5p miRNA overexpression and inhibition. HAoSMCs were treated with combinations of the inflammatory cytokine tumor necrosis factor-α (TNF-α), and atorvastatin or lovastatin. HAoSMC Bulk RNA-seq transcriptomics of HAoSMCs revealed downstream regulatory networks modulated by let-7d-5p miRNA overexpression and statins. Proteome profiler cytokine array, Western blotting and quantitative PCR analyses were performed on HAoSMCs to validate key findings. ResultsLet-7d-5p overexpression significantly attenuated TNF-α-induced upregulation of IL-6, ICAM1, VCAM1, CCL2, CD68, MYOCD gene expression in HAoSMCs (p<0.05). Statins (atorvastatin, lovastatin) significantly attenuated inflammatory gene expression and upregulated Let-7d levels in HAoSMCs (p<0.05). Bulk RNA-seq analysis of a dual Let-7d-5p overexpression/statin therapy in HAoSMCs revealed that let-7d-5p activation and statins converge on key inflammatory pathways (IL-6, IL-1β, TNF-α, IFN-γ). Let-7d-5p overexpression led to reduced expression of the ox-LDL receptor OLR1, and this was associated with lower ox-LDL uptake in HAoSMCs. In silico analysis of smooth muscle cell phenotypic switching shows that overexpression of let-7d-5p in HAoSMCs maintains a contractile phenotype. ConclusionsTargeting the Let-7 network alongside statins can modulate HAoSMC activation and attenuate key inflammatory pathway signals.

Utility of intracranial EEG networks depends on re-referencing and connectivity choice.

Studies of intracranial EEG networks have been used to reveal seizure generators in patients with drug-resistant epilepsy. Intracranial EEG is implanted to capture the epileptic network, the collection of brain tissue that forms a substrate for seizures to start and spread. Interictal intracranial EEG measures brain activity at baseline, and networks computed during this state can reveal aberrant brain tissue without requiring seizure recordings. Intracranial EEG network analyses require choosing a reference and applying statistical measures of functional connectivity. Approaches to these technical choices vary widely across studies, and the impact of these technical choices on downstream analyses is poorly understood. Our objective was to examine the effects of different re-referencing and connectivity approaches on connectivity results and on the ability to lateralize the seizure onset zone in patients with drug-resistant epilepsy. We applied 48 pre-processing pipelines to a cohort of 125 patients with drug-resistant epilepsy recorded with interictal intracranial EEG across two epilepsy centres to generate intracranial EEG functional connectivity networks. Twenty-four functional connectivity measures across time and frequency domains were applied in combination with common average re-referencing or bipolar re-referencing. We applied an unsupervised clustering algorithm to identify groups of pre-processing pipelines. We subjected each pre-processing approach to three quality tests: (i) the introduction of spurious correlations; (ii) robustness to incomplete spatial sampling; and (iii) the ability to lateralize the clinician-defined seizure onset zone. Three groups of similar pre-processing pipelines emerged: common average re-referencing pipelines, bipolar re-referencing pipelines and relative entropy-based connectivity pipelines. Relative entropy and common average re-referencing networks were more robust to incomplete electrode sampling than bipolar re-referencing and other connectivity methods (Friedman test, Dunn-Šidák test P < 0.0001). Bipolar re-referencing reduced spurious correlations at non-adjacent channels better than common average re-referencing (Δ mean from machine ref = -0.36 versus -0.22) and worse in adjacent channels (Δ mean from machine ref = -0.14 versus -0.40). Relative entropy-based network measures lateralized the seizure onset hemisphere better than other measures in patients with temporal lobe epilepsy (Benjamini-Hochberg-corrected P < 0.05, Cohen's d: 0.60-0.76). Finally, we present an interface where users can rapidly evaluate intracranial EEG pre-processing choices to select the optimal pre-processing methods tailored to specific research questions. The choice of pre-processing methods affects downstream network analyses. Choosing a single method among highly correlated approaches can reduce redundancy in processing. Relative entropy outperforms other connectivity methods in multiple quality tests. We present a method and interface for researchers to optimize their pre-processing methods for deriving intracranial EEG brain networks.

PagWOX11/12a from hybrid poplar enhances drought tolerance by modulating reactive oxygen species

WOX11/12 is a homeobox gene of WOX11 and WOX12 in Arabidopsis that plays important roles in crown root development and growth. It has been reported that WOX11/12 participates in adventitious root (AR) formation and different abiotic stress responses, but the downstream regulatory network of WOX11/12 in poplar remains to be further investigated. In this study, we found that PagWOX11/12a is strongly induced by PEG-simulated drought stress. PagWOX11/12a-overexpressing poplar plantlets showed lower oxidative damage levels, greater antioxidant enzyme activities and reactive oxygen species (ROS) scavenging capacity than non-transgenic poplar plants, whereas PagWOX11/12a dominant repression weakened root biomass accumulation and drought tolerance in poplar. RNA-seq analysis revealed that several differentially expressed genes (DEGs) regulated by PagWOX11/12a are involved in redox metabolism and drought stress response. We used RT-qPCR and yeast one-hybrid (Y1H) assays to validate the downstream target genes of PagWOX11/12a. These results provide new insights into the biological function and molecular regulatory mechanism of WOX11/12 in the abiotic resistance processes of poplar.

Multi-resolution sequence and structure feature extraction for binding site prediction

Circular ribonucleic acids (circRNAs) are single-stranded RNA molecules that form loops and are widely expressed in various cells and tissues. They interact with RNA-binding proteins (RBPs) and play a vital regulatory role in the onset and development of several diseases. Researchers have proposed various hybrid architecture prediction methods based on convolutional neural networks and recurrent neural networks to recognize the interactions and sites between circRNAs and RBPs and thus reveal the biological functions of circRNAs. However, existing methods usually ignore the structural information of circRNA, which may affect the modeling of circRNA and RBP binding modes. To address these problems, we propose a prediction model based on multi-resolution feature extraction. First, it generates sequence features using unsupervised word embedding and nucleotide density. Then, it uses implicit and explicit pseudo-secondary structure hybrid encoding to fuse sequence and structural information and better simulate circRNA-RBP binding patterns. Second, it uses an enhanced bidirectional sample convolution and interaction network encoder to capture and integrate high-order features of distinct resolutions from the multi-scale convolution module. This provides rich semantic input to the downstream bidirectional long short-term memory network to improve prediction accuracy. Experimental results on 37 circRNA and 31 linear RNA datasets show that our method has significant advantages in identifying RNA-RBP interactions. Furthermore, the four motifs learned by our method are verified against existing motif databases, indicating that it can discover biologically meaningful circRNA-RBP binding patterns.

The role of language-related functional brain regions and white matter tracts in network plasticity of post-stroke aphasia.

The neural mechanisms underlying language recovery after a stroke remain controversial. This review aimed to summarize the plasticity and reorganization mechanisms of the language network through neuroimaging studies. Initially, we discussed the involvement of right language homologues, perilesional tissue, and domain-general networks. Subsequently, we summarized the white matter functional mapping and remodeling mechanisms associated with language subskills. Finally, we explored how non-invasive brain stimulation (NIBS) promoted language recovery by inducing neural network plasticity. It was observed that the recruitment of right hemisphere language area homologues played a pivotal role in the early stages of frontal post-stroke aphasia (PSA), particularly in patients with larger lesions. Perilesional plasticity correlated with improved speech performance and prognosis. The domain-general networks could respond to increased "effort" in a task-dependent manner from the top-down when the downstream language network was impaired. Fluency, repetition, comprehension, naming, and reading skills exhibited overlapping and unique dual-pathway functional mapping models. In the acute phase, the structural remodeling of white matter tracts became challenging, with recovery predominantly dependent on cortical activation. Similar to the pattern of cortical activation, during the subacute and chronic phases, improvements in language functions depended, respectively, on the remodeling of right white matter tracts and the restoration of left-lateralized language structural network patterns. Moreover, the midline superior frontal gyrus/dorsal anterior cingulate cortex emerged as a promising target for NIBS. These findings offered theoretical insights for the early personalized treatment of aphasia after stroke.

Experimental demonstration of quantum downstream access network in continuous variable quantum key distribution with a local local oscillator

Experimental demonstration of quantum downstream access network in continuous variable quantum key distribution with a local local oscillator

Revisiting Disentanglement in Downstream Tasks: A Study on Its Necessity for Abstract Visual Reasoning

In representation learning, a disentangled representation is highly desirable as it encodes generative factors of data in a separable and compact pattern. Researchers have advocated leveraging disentangled representations to complete downstream tasks with encouraging empirical evidence. This paper further investigates the necessity of disentangled representation in downstream applications. Specifically, we show that dimension-wise disentangled representations are unnecessary on a fundamental downstream task, abstract visual reasoning. We provide extensive empirical evidence against the necessity of disentanglement, covering multiple datasets, representation learning methods, and downstream network architectures. Furthermore, our findings suggest that the informativeness of representations is a better indicator of downstream performance than disentanglement. Finally, the positive correlation between informativeness and disentanglement explains the claimed usefulness of disentangled representations in previous works. The source code is available at https://github.com/Richard-coder-Nai/disentanglement-lib-necessity.git

Abstract 2691: Mapping the landscape of innate immune pattern recognition receptors in pancreatic cancer

Abstract Pancreatic ductal adenocarcinoma (PDAC) accounts for 90% of all pancreatic cancer cases, and is among the most aggressive and lethal malignancies worldwide. PDAC is driven by genetic alterations in the pancreatic epithelium (e.g. KRAS, TRP53) coupled with a dysregulated innate immune response, the latter leading to an inflammatory tumor microenvironment enriched in innate immune cells (e.g. macrophages). However, immune-based treatment regimens for PDAC patients have primarily focused on immunotherapy with adaptive immune checkpoint inhibitors (e.g. PD-1) which disappointingly have yielded minimal clinical benefit. Thus, there is an unmet clinical need to identity disease-associated innate immune regulators as therapeutic targets in PDAC. Innate immune responses depend on a series of cell surface, endosomal and cytosolic pattern recognition receptors (PRRs) that are expressed in immune and non-immune cells. PRRs are classified into several structurally and functionally conserved subfamilies, including Toll-like receptors (TLRs), Absent in melanoma 2 (AIM2)-like receptors (ALRs), and Nod-like receptors (NLRs). A subset of NLRs and AIM2 are also well documented for their formation of multiprotein complexes called ‘inflammasomes’ - comprising the adaptor ASC and Caspase-1 - which direct the maturation and release of bioactive pro-inflammatory cytokines, IL-18 and IL-1β. However, the role of these critical innate immune regulators in PDAC has been underexplored, and is virtually unknown. By coupling endoscopic ultrasound-guided fine needle aspiration (EUS FNA) to capture tumor biopsies from all stages of PDAC, with whole transcriptome profiling of PDAC patient primary tumors, we reveal specific enrichment of the innate immune TLR2 and ASC inflammasome molecular pathways. Augmented TLR2 and ASC expression was prognostic for low survival, and a “TLR2 activation” signature was also predictive for chemoresistance. Antibody-mediated anti-TLR2 therapy suppressed tumor growth in human PDAC tumor xenografts (patient and cell line) in both immunocompromised and ‘humanized’ mice. With respect to ASC, using the genetically engineered KPC PDAC mouse model, we show that Caspase-1 activation, a read-out for inflammasome activation, is upregulated in pancreatic tumors. Moreover, genetic ablation of ASC in KPC mice markedly suppressed the development of pancreatic tumors, which was associated with significant reductions in levels of both Caspase-1 activation and inflammation. Transcriptome profiling further identified unique downstream molecular networks associated with a range of oncogenic cellular processes. Taken together, our results support innate immune TLR2- and ASC inflammasome-based therapeutic targeting in PDAC, with further clinical utility that ASC inflammasome and TLR2 activation is prognostic, with TLR2 also predictive for chemoresponsiveness. Citation Format: Bassam Kashgari, Louise McLeod, Linden J. Gearing, Daniel Croagh, Brendan J. Jenkins. Mapping the landscape of innate immune pattern recognition receptors in pancreatic cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2691.

A semantic-driven coupled network for infrared and visible image fusion

In order to be adapted to high-level vision tasks, several infrared and visible image fusion methods cascade with the downstream network to enhance the semantic information of fusion results. However, due to the feature-level heterogeneities between fusion and downstream tasks, these methods suffer from the loss of pixel-level information and incomplete reconstruction of semantic-level information. To further improve the performance of fusion images in high-level vision tasks, we propose a semantic-driven coupled network for infrared and visible image fusion, terms as SDCFusion. Firstly, to address feature heterogeneity, we couple the segmentation and fusion networks into a joint framework such that both networks share the multi-level cross-modality coupled features. Based on the joint optimization of dual tasks, a joint action between fusion and downstream tasks is formed to force the cross-modality coupled features modeled on both pixel domain and semantic domain. Subsequently, to guide the semantic information reconstruction, we cascade two networks to form the semantic-based driven action, which continuously optimizes the fusion image to achieve semantic representation capacity. In addition, we introduce an adaptive training strategy to reduce the complexity of dual-task training. Specifically, an mIoU-based semantic measurement weight is designed to balance the joint action and driven action throughout the training process. We evaluate our method at both pixel information and semantic information levels, respectively. The qualitative and quantitative experiments verify the superiority of SDCFusion in terms of visual effects and metrics. The object detection and semantic segmentation experiments demonstrate that SDCFusion achieves superior performance in high-level vision tasks. The source code is available at https://github.com/XiaoW-Liu/SDCFusion.

Subtracting-adding strategy for necrotic lesion segmentation in osteonecrosis of the femoral head.

Osteonecrosis of the femoral head (ONFH) is a severe bone disease that can progressively lead to hip dysfunction. Accurately segmenting the necrotic lesion helps in diagnosing and treating ONFH. This paper aims at enhancing deep learning models for necrosis segmentation. Necrotic lesions of ONFH are confined to the femoral head. Considering this domain knowledge, we introduce a preprocessing procedure, termed the "subtracting-adding" strategy, which explicitly incorporates this domain knowledge into the downstream deep neural network input. This strategy first removes the voxels outside the predefined volume of interest to "subtract" irrelevant information, and then it concatenates the bone mask with raw data to "add" anatomical structure information. Each of the tested off-the-shelf networks performed better with the help of the "subtracting-adding" strategy. The dice similarity coefficients increased by 10.93%, 9.23%, 9.38% and 1.60% for FCN, HRNet, SegNet and UNet, respectively. The improvements in FCN and HRNet were statistically significant. The "subtracting-adding" strategy enhances the performance of general-purpose networks in necrotic lesion segmentation. This strategy is compatible with various semantic segmentation networks, alleviating the need to design task-specific models.

자동차 산업 공급 네트워크의 상류와 하류 구성 동인의 비교

Modern industry is a competition between supply networks. Knowing and managing the supply network well is an essential survival requirement for a company. There is a difference in orientation between the upstream and downstream of the supply network, which causes structural differences between the upstream and downstream. This study aims to reveal the differences in the configuration drivers of the upstream and downstream automotive supply networks. Using transaction data from the automotive industry, we compared the drivers that constitute supply networks and create upstream and downstream links. The Exponential Random Graph Models (ERGMs) were used to evaluate the significance of configuration drivers, including company attributes, dyadic attributes, and structural characteristics. As a result, the age of supplier, size of the purchaser, sales of the purchaser, asset of the purchaser, transitivity, and activity were found to be common drivers of upstream and downstream. On the other hand, the age of the purchaser, the difference in sales between the two companies, and popularity were found to be significant drivers only in the downstream network. These research results are the first to empirically compare the upstream and downstream supply networks, and have academic contribution as the first step in finding the cause of the differences in network structure between the upstream and downstream networks.

Bioinformatics analysis and experimental validation of key genes associated with lumbar disc degeneration and biomechanics

BackgroundLumbar disc degeneration (LDD) is an important pathological basis for the development of degenerative diseases of the lumbar spine. Most clinical patients have low back pain as their main symptom. The deterioration of the biomechanical environment is an important cause of LDD. Although there is a large amount of basic research on LDD, there are fewer reports that correlate biomechanical mechanisms with basic research. Our research aims to identify 304 key genes involved in LDD due to biomechanical deterioration, using a bioinformatics approach. We focus on SMAD3, CAV1, SMAD7, TGFB1 as hub genes, and screen for 30 potential target drugs, offering novel insights into LDD pathology and treatment options. MethodsThe Gene Cards, GenCLip3, OMIM and Drugbank databases were explored to obtain genes associated with biomechanics and LDD, followed by making veen plots to obtain both co-expressed genes. GO enrichment analysis and KEGG pathway analysis of the co-expressed genes were obtained using the DAVID online platform and visualised via a free online website. Protein interaction networks (PPI) were obtained through the STRING platform and visualised through Cytoscape 3.9.0. These genes were predicted for downstream interaction networks using the STITCH platform. Then, the GSE56081 dataset was used to validate the key genes. RT-PCR was used to detect mRNA expression of core genes in the degenerated nucleus pulposus (NP) samples and western bolt was used for protein expression. Lastly, the obtained hub genes were searched in the drug database (DGIdb) to find relevant drug candidates. ResultsFrom the perspective of biomechanics-induced LDD, we obtained a total of 304 genes, the GO functional enrichment and KEGG pathway enrichment analysis showed that the functions of these genes are mostly related to inflammation and apoptosis. The PPI network was constructed and four Hub genes were obtained through the plug-in of Cytoscape software, namely SMAD3, CAV1, SMAD7 and TGFB1. The analysis of key genes revealed that biomechanical involvement in LDD may be related to the TGF-β signaling pathway. Validation of the GSE56081 dataset revealed that SMAD3 and TGFB1 were highly expressed in degenerating NP samples. RT-PCR results showed that the mRNA expression of SMAD3 and TGFB1 was significantly increased in the severe degeneration group; Western blot results also showed that the protein expression of TGFB1 and P-SMAD3 was significantly increased. In addition, we identified 30 potential drugs. ConclusionThis study presented a new approach to investigate the correlation between biomechanical mechanisms and LDD. The deterioration of the biomechanical environment may cause LDD through the TGF-β signaling pathway. TGFB1 and SMAD3 are important core targets. The important genes, pathways and drugs obtained in this study provided a new basis and direction for the study, diagnosis and treatment of LDD.