Metabolic network analysis of pre-ASD newborns and 5-year-old children with autism spectrum disorder.

Classical metabolomic and new metabolic network methods were used to study the developmental features of autism spectrum disorder (ASD) in newborns (n = 205) and 5-year-old children (n = 53). Eighty percent of the metabolic impact in ASD was caused by 14 shared biochemical pathways that led to decreased anti-inflammatory and antioxidant defenses, and to increased physiologic stress molecules like lactate, glycerol, cholesterol, and ceramides. CIRCOS plots and a new metabolic network parameter, V ° net, revealed differences in both the kind and degree of network connectivity. Of 50 biochemical pathways and 450 polar and lipid metabolites examined, the developmental regulation of the purine network was most changed. Purine network hub analysis revealed a 17-fold reversal in typically developing children. This purine network reversal did not occur in ASD. These results revealed previously unknown metabolic phenotypes, identified new developmental states of the metabolic correlation network, and underscored the role of mitochondrial functional changes, purine metabolism, and purinergic signaling in autism spectrum disorder.

TDG prediction model improvement by analysis and validation of experiments on a dam model.

The combination of aerated flows and a high-pressure environment in a stilling basin can result in the supersaturation of total dissolved gas (TDG) downstream of hydraulic projects, posing an ecological risk to aquatic populations by inducing gas bubble disease (GBD) or other negative effects. There is limited literature reporting TDG mass transfer experiments on a complete physical dam model; most existing research is based on measurements in prototype tailwaters. In this study, TDG mass transfer experiments were conducted on a physical model of an under-constructed dam, with TDG-supersaturated water as the inflow, and TDG concentrations were meticulously monitored within the stilling basin. The measurements indicate that the TDG saturation at the outlet of the stilling basin decreased by 13.7% and 10.6% compared to the inlet for the two cases, respectively. Subsequently, an improved TDG prediction model was developed by incorporating a sub-grid air entrainment model and a phase-constrained scalar model. The numerical simulation results were compared with experimental data, indicating a maximum error in TDG saturation at all measured points of less than ± 3%. Moreover, the TDG saturation showed an error of only ± 0.3% at the outlet of the stilling basin. This model has broad applicability to various flow types for obtaining TDG mass transfer results and evaluating mitigation measures of TDG supersaturation to reduce the harmful effects on aquatic ecosystems.

Estimating the water quality index based on interpretable machine learning models.

The water quality index (WQI) is an important tool for evaluating the water quality status of lakes. In this study, we used the WQI to evaluate the spatial water quality characteristics of Dianchi Lake. However, the WQI calculation is time-consuming, and machine learning models exhibit significant advantages in terms of timeliness and nonlinear data fitting. We used a machine learning model with optimized parameters to predict the WQI, and the light gradient boosting machine achieved good predictive performance. The machine learning model trained based on the entire Dianchi Lake water quality data achieved coefficient of determination (R2), mean square error, and mean absolute error values of 0.989, 0.228, and 0.298, respectively. In addition, we used the Shapley additive explanations (SHAP) method to interpret and analyse the machine learning model and identified the main water quality parameter that affects the WQI of Dianchi Lake as NH4+-N. Within the entire range of Dianchi Lake, the SHAP values of NH4+-N varied from -9 to 3. Thus, in future water environmental governance, it is necessary to focus on NH4+-N changes. These results can provide a reference for the treatment of lake water environments.

A design of a nature-like fishway to solve the fractured river connectivity caused by small hydropower based on hydrodynamics and fish behaviors.

Achieving a harmonious alignment between the biological characteristics of fish and hydrodynamics patterns is crucial for ensuring the efficacy of fish passage facilities. In this study, based on the hydrodynamic characteristics of the river and the biological characteristics of fish, we evaluated the internal flow field in the nature-like fishway of Congen II hydropower station located along the Chabao river and explored methods to improve the operation efficiency. Based on comprehensive considerations of the flow field, turbulent kinetic energy, and the migration pathways of fish, it is found that the implementation of a continuous oblique bottom slope represents a more cost-effective and operationally convenient solution. The influence of different permutation of bulkheads in the nature-like fishway on operational efficiency was further examined. Our investigation revealed that the nature-like fishway with the continuous slope of 2% and the arrangement of three bulkheads in each row (model 3) exhibited a relatively simple velocity distribution and linear flow line, which poses challenges for fish in locating resting areas. In addition, the distribution of low turbulence kinetic energy area in the mainstream made it less favorable for fish to transition from the mainstream to the rest area within the fishway. The nature-like fishway with the continuous slope of 2% and the arrangement of two or three bulkheads in staggered rows (model 4) demonstrated better performance. Several potential fish migration routes for both model 3 and model 4 were proposed based on the numerical simulation results. In model 3, fish exhibited a continuous sprint through the concentrated high-speed area, which was less favorable for fish to rest and forage. In contrast, model 4 exhibited a diversified flow velocity distribution, enabling fish to make timely changes in their direction during migration. This feather proved to be advantageous in enhancing fish migration within the passage. The design of nature-like fishway in this study provides an important reference and technical support for the construction and optimization of the nature-like fishway for low dams, and is of great significance for restoring river connectivity destroyed by small hydropower construction and improving fish migration.

Conflict between urbanization and water environmental protection: Lessons from the Xiangjiang River Basin in China.

China, the largest developing country, has experienced rapid urbanization since its reform and opening-up. However, the increasing pollution load from urban areas has deteriorated urban river water quality, contradicting the concept of sustainable and green development promoted by the Chinese government. This situation elucidates governmental shortcomings in systematic environmental protection. Our study revealed that the current wastewater treatment plant (WWTP) discharge standards in urban areas are insufficient for attaining the desired urban river water quality and thus intensify the conflict between urbanization and water environmental protection. As urbanization continues, the urban population will grow, further exacerbating pollution and conflict. Our focus was the Xiangjiang River basin in Zunyi, a typical urbanized city in China. Using a validated one-dimensional mathematical model, we compared the water quality in the Xiangjiang River between current and upgraded WWTP discharge standards. The results showed that the water quality in the Xiangjiang River falls short of the standards, with more than 60 % of the river exceeding limits. However, upgrading WWTP discharge standards significantly reduces the proportion of river sections exceeding limits, with only 0.4 % exceeding standards during specific periods. This enhancement greatly improved the Xiangjiang River's water quality, aided in restoring the entire water environment in the basin, and supported water environmental protection goals. Our research findings offer crucial support for local governments in shaping comprehensive water environmental protection policies and insights for addressing similar environmental challenges caused by rapid urbanization in other developing regions.

Neuroimaging features in a patient with non-ketotic hyperglycaemic seizures: A case report.

BACKGROUND: Non-ketotic hyperglycaemic (NKH) seizures are a rare neurological complication of diabetes caused by hyperglycaemia in non-ketotic and non-hyperosmotic states. The clinical characteristics of NKH seizures are atypical and lack unified diagnostic criteria, leading to potential misdiagnoses in the early stages of the disease. CASE SUMMARY: This report presents a rare case of NKH seizures in a 52-year-old male patient with a history of type 2 diabetes mellitus. We performed comprehensive magnetic resonance imaging (MRI) studies at admission, 12 d post-admission, and 20 d post-discharge. The imaging techniques included contrast-enhanced head MRI, T2-weighted imaging (T2WI), fluid-attenuated inversion recovery (FLAIR), diffusion-weighted imaging, susceptibility-weighted imaging, magnetic resonance spectroscopy (MRS), and magnetic resonance venography. At the time of admission, T2WI and FLAIR of the cranial MRI showed that the left parieto-occipital cortex had gyrus-like swelling and high signal, and subcortical stripes had low signal. MRS showed a reduced N-acetylaspartate peak and increased creatine and choline peaks in the affected areas. A follow-up MRI 20 d later showed that the swelling and high signal of the left parieto-occipital cortex had disappeared, and the low signal of the subcortex had disappeared. CONCLUSION: This case study provides valuable insights into the potential pathogenesis, diagnosis, and treatment of NKH seizures. The comprehensive MRI findings highlight the potential utility of various MRI sequences in diagnosing and characterizing NKH seizures.

Lightweight Vehicle Detection Based on Improved YOLOv5s.

A vehicle detection algorithm is an indispensable component of intelligent traffic management and control systems, influencing the efficiency and functionality of the system. In this paper, we propose a lightweight improvement method for the YOLOv5 algorithm based on integrated perceptual attention, with few parameters and high detection accuracy. First, we propose a lightweight module IPA with a Transformer encoder based on integrated perceptual attention, which leads to a reduction in the number of parameters while capturing global dependencies for richer contextual information. Second, we propose a lightweight and efficient multiscale spatial channel reconstruction (MSCCR) module that does not increase parameter and computational complexity and facilitates representative feature learning. Finally, we incorporate the IPA module and the MSCCR module into the YOLOv5s backbone network to reduce model parameters and improve accuracy. The test results show that, compared with the original model, the model parameters decrease by about 9%, the average accuracy (mAP@50) increases by 3.1%, and the FLOPS does not increase.

Metabolic Connectome and Its Role in the Prediction, Diagnosis, and Treatment of Complex Diseases.

The interconnectivity of advanced biological systems is essential for their proper functioning. In modern connectomics, biological entities such as proteins, genes, RNA, DNA, and metabolites are often represented as nodes, while the physical, biochemical, or functional interactions between them are represented as edges. Among these entities, metabolites are particularly significant as they exhibit a closer relationship to an organism's phenotype compared to genes or proteins. Moreover, the metabolome has the ability to amplify small proteomic and transcriptomic changes, even those from minor genomic changes. Metabolic networks, which consist of complex systems comprising hundreds of metabolites and their interactions, play a critical role in biological research by mediating energy conversion and chemical reactions within cells. This review provides an introduction to common metabolic network models and their construction methods. It also explores the diverse applications of metabolic networks in elucidating disease mechanisms, predicting and diagnosing diseases, and facilitating drug development. Additionally, it discusses potential future directions for research in metabolic networks. Ultimately, this review serves as a valuable reference for researchers interested in metabolic network modeling, analysis, and their applications.

TDG generation in a ski-jump spillway with a fully or partial-flip bucket.

Ski-jump spillways are frequently used as discharge structures for high dams during floods with high energy heads. The selection of bucket types at the end of spillways has a pronounced effect on the hydraulics of jet characteristics, such as trajectories and entrained air features. However, there is no literature reporting how changes in the bucket types influence TDG generation. This study compares the hydraulic characteristics and TDG mass transfer properties of a hydraulic project under construction using both the traditional fully-flip bucket and the partial-flip bucket configurations. The results indicate that, the use of the partial-flip bucket at the end of the spillway significantly disperses the water flow and yields better energy dissipation effects. At low flow rates (lower than 400 m3/s for the dam in this study), there is little difference in the downstream TDG saturation between the traditional fully-flip bucket and the partial-flip bucket, the average difference is 1.6 % in three cases with a low flow rate. However, at high flow rates (higher than 400 m3/s), the partial-flip bucket generates more TDG compared to the traditional fully-flip bucket, reaching up to 6.2 % at the maximum flow rate. This phenomenon stems from significant changes in the hydrodynamics of the stilling basin at high flow rates due to variations in the flip bucket type. When strict control of TDG generation is necessary downstream of dams, the use of the partial-flip bucket should be carefully considered. This is because, at high flow rates, the partial-flip bucket might result in higher TDG saturation than the fully-flip bucket.

MetDIT: Transforming and Analyzing Clinical Metabolomics Data with Convolutional Neural Networks.

Clinical metabolomics is growing as an essential tool for precision medicine. However, classical machine learning algorithms struggle to comprehensively encode and analyze the metabolomics data due to their high dimensionality and complex intercorrelations. This article introduces a new method called MetDIT, designed to analyze intricate metabolomics data effectively using deep convolutional neural networks (CNN). MetDIT comprises two components: TransOmics and NetOmics. Since CNN models have difficulty in processing one-dimensional (1D) sequence data efficiently, we developed TransOmics, a framework that transforms sequence data into two-dimensional (2D) images while maintaining a one-to-one correspondence between the sequences and images. NetOmics, the second component, leverages a CNN architecture to extract more discriminative representations from the transformed samples. To overcome the overfitting due to the small sample size and class imbalance, we introduced a feature augmentation module (FAM) and a loss function to improve the model performance. Furthermore, we systematically optimized the model backbone and image resolution to balance the model parameters and computational costs. To demonstrate the performance of the proposed MetDIT, we conducted extensive experiments using three different clinical metabolomics data sets and achieved better classification performance than classical machine learning methods used in metabolomics, including Random Forest, SVM, XGBoost, and LightGBM. The source code is available at the GitHub repository at https://github.com/Li-OmicsLab/MetDIT, and the WebApp can be found at http://metdit.bioinformatics.vip/.

Determination of chloropropanol esters and glycidyl esters in instant noodles based on solid-phase microextraction with chitosan-ß-cyclodextrin coated fiber.

We developed a method for the determination of chloropropanol esters and glycidyl esters (GE) in instant noodles using solid-phase microextraction with chitosan-ß-cyclodextrin (CS-ß-CD) coated fiber coupled with gas chromatography-tandem mass spectrometry. The developed low-cost fiber coating can improve the sensitivity of the method. Immobilized enzymes can improve operational stability and reusability compared to free enzymes, thereby reducing costs. The adsorption isotherm was modeled using the Langmuir model, while the adsorption kinetics followed the pseudo second-order model. The limit of detection was 0.3 ng/L. The method exhibited satisfactory recoveries for the analytes, ranging from 80.2 % to 105.3 %, with relative standard deviations < 9.9 %. Furthermore, the results of the exposure assessment showed that chloropropanol esters do not pose unacceptable risks to different age groups. However, the margin of exposure for GE suggested a potential health risk for populations between the ages of 3 and 12 years old.

The Prevalence, Epidemiological, and Molecular Characterization of Methicillin-Resistant Staphylococcus aureus (MRSA) in Macau (2017-2022).

Macau, recognized as a global tourism hub and the world's most densely populated region, provides a unique environment conducive to methicillin-resistant Staphylococcus aureus (MRSA) transmission in healthcare and community settings, posing a significant public health concern both locally and globally. The epidemiology and molecular characteristics of MRSA in the distinct city of Macau remain largely unelucidated. This five-year longitudinal study (2017-2022) examined the local prevalence and molecular typing of MRSA in Macau, with future MRSA type distribution predicted through ARIMA modeling. We subsequently analyzed the epidemiological characteristics of MRSA, including specimen source, clinical department, collection year, season, patient age, sex, and the annual number of tourists. Comprehensive antibiotic resistance profiles of the strains were also assessed. Of 504 clinically isolated S. aureus strains, 183 (36.3%) were identified as MRSA by the cefoxitin disk diffusion method and validated through multi-locus sequence typing (MLST). The MRSA detection rate showed an upward trend, increasing from 30.1% in 2017 to 45.7% in 2022. SCCmec type IV was predominant (28.9%), followed by types II (25.4%), III (22.1%), and V (22.1%). The primary sources of MRSA isolates were sputum (39.2%) and secretions (25.6%). Older age emerged as a risk factor for MRSA infection, whereas no significant associations were found with seasonal variations, gender, or the annual number of tourists. Despite displaying universal resistance to cefoxitin, oxacillin, and benzylpenicillin, MRSA isolates in Macau remained fully sensitive to vancomycin, tigecycline, quinupristin, nitrofurantoin, and linezolid. Continuous surveillance and analysis of MRSA distribution in Macau could provide invaluable insights for the effective management of MRSA prevention and control measures within healthcare settings.

Sports game intervention aids executive function enhancement in children with autism - An fNIRS study.

Executive dysfunction is a prevalent issue in children diagnosed with autism spectrum disorder (ASD). While the efficacy of physical exercise in enhancing cognitive abilities in these children is well-documented, research exploring the relationship between physical exercise and brain function remains limited. This study aimed to investigate the impact of cognitively stimulating exercise on executive functions (EF) in children with ASD. The study enrolled thirty children with ASD who were randomly allocated into two groups: a sports game learning group (n = 15) and a control group (n = 15). Functional near-infrared spectroscopy was utilized to monitor cerebral function alterations pre- and post- an eight-week intervention program. The study focused on three core components of executive function: working memory, inhibitory control (IC), and cognitive flexibility (CF). Results revealed a significant improvement in the EF in the intervention group. After eight weeks of intervention, neural activity, along with improved EF performance, was enhanced significantly in the prefrontal cortex (PFC). During post-intervention, EF tasks were also significantly activated in the dorsolateral PFC, orbitofrontal cortex, and frontal pole area. Furthermore, an increase in short-distance functional connectivity within the PFC was observed during resting states. These results imply that engagement in sports game training can significantly improve EF information processing, augmenting task-related cortical activations and the efficiency of brain function networks in children with ASD.

Climate change: approach to intervention using expression vector for carbonic anhydrase via glycosylphosphatidylinositol.

Climate change is a change in the usual weather found in a place. The climate change has a major impact not only on natural disasters of the Earth but also on human health. The climate crisis is then no longer a future concern. It includes both the global warming driven by human emissions of greenhouse gases (GHG), and the resulting large-scale shifts in weather patterns. Global warming can occur from a variety of causes, both natural and human induced. The primary GHG in Earth's atmosphere, listed in decreasing order of average global mole fraction, are: water vapor (H2O), carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O), and ozone (O3). Today, scientists around the world continue to try and solve the puzzle of climate change. It is clear that to address climate change, the amount of CO2 released into the atmosphere by industrial process has to be reduced because once it is added to the atmosphere, it can continue to affect climate for thousands of years. For such a purpose, an approach to intervention using expression vectors for any protein targeting to the cell plasma membrane via the glycosylphosphatidylinositol, GPI, anchor is suggested. The resulting GPI-anchored proteins would be useful for studying intermolecular interactions, especially gene-environment interactions, in investigating the potential impact of any chemical compounds on any genes of interest and could be used for carbonic anhydrase (CA)-based CO2-capture (environmental application). This approach would be crucial not only for capturing CO2 via GPI and CA but also for the production of CA enzyme as well as its stabilization and therefore useful for combating the global warming of climate change.

Determination of parabens in breast milk using stir bar sorptive extraction coupled with UHPLC-UV.

We developed an analytical method based on ultra-high performance liquid chromatography with UV detection, using a stir bar coated with amino/hydroxyl bifunctional microporous organic network (B-MON), for the analysis of parabens in breast milk samples. B-MON demonstrated superior performance with maximal methylparaben adsorption of 112.15 mg/g. Kinetic fitting revealed that outer diffusion was the key limiting step, and the adsorption was chemisorption. The thermodynamic analysis demonstrated that increased methylparaben adsorption was found at higher temperatures in spontaneous processes. The developed approach showed excellent linearity (R2 ≥ 0.9964) and a low detection limit (0.01 µg/L). Recoveries ranged from 85.8 to 105.5 % and the relative standard deviation was lower than 9.2 %. Based on the daily exposure assessment, these pollutants do not pose unacceptable health hazards to babies. However, the high detection frequencies (41.9%-93.5 %) suggest that breast milk still should be monitored.

Effect of total dissolved gas supersaturation and flow velocity on survival and swimming ability of juvenile Schizothorax prenanti.

Although developing large-scale hydropower cascades in the upper Yangtze River effectively improves the hydropower resource utilization, it produces total dissolved gas (TDG) supersaturation. In the flood season, the high level of TDG supersaturation (TDGS) frequently occurs in the downstream of dams, causing migratory fish to suffer from gas bubble trauma (GBT) and reducing their survival and swimming ability. Currently, there is a deficiency in particular approaches to evaluate the ecological hazard posed by TDGS on migratory fish as they traverse different flow velocities within their migratory routes. This study assessed the vulnerability of juvenile Schizothorax prenanti (S. prenanti) to GBT from the static setting to 9.0 BL/s during exposure to nominal levels of 100%, 110%, 120% and 130% TDG. The mortality occurs when the flow velocity surpasses 6.0 and 7.5 BL/s in 100% and 110% TDG levels, respectively. For fish exposed to 120% and 130% TDG levels, the relationship between survival time and flow velocity is an approximately inverse bell-shaped curve with increasing velocity. The optimal velocity of maximal survival time of juvenile S. prenanti is 3.0 and 4.5 BL/s in 120% and 130% TDG-supersaturated water. Both TDG level and flow velocity significantly affect burst swimming speed (Uburst) and critical swimming speed (Ucrit). The cases involving GBT showed substantial declines in Uburst and Ucrit, exceeding 6.0 BL/s and TDG levels greater than 120%. The results may contribute to formulating a specific management strategy for hydropower operation during the migratory period and conserving vulnerable species in the Yangtze River.

Metabolic features of treatment-refractory major depressive disorder with suicidal ideation.

Peripheral blood metabolomics was used to gain chemical insight into the biology of treatment-refractory Major Depressive Disorder with suicidal ideation, and to identify individualized differences for personalized care. The study cohort consisted of 99 patients with treatment-refractory major depressive disorder and suicidal ideation (trMDD-SI n = 52 females and 47 males) and 94 age- and sex-matched healthy controls (n = 48 females and 46 males). The median age was 29 years (IQR 22-42). Targeted, broad-spectrum metabolomics measured 448 metabolites. Fibroblast growth factor 21 (FGF21) and growth differentiation factor 15 (GDF15) were measured as biomarkers of mitochondrial dysfunction. The diagnostic accuracy of plasma metabolomics was over 90% (95%CI: 0.80-1.0) by area under the receiver operator characteristic (AUROC) curve analysis. Over 55% of the metabolic impact in males and 75% in females came from abnormalities in lipids. Modified purines and pyrimidines from tRNA, rRNA, and mRNA turnover were increased in the trMDD-SI group. FGF21 was increased in both males and females. Increased lactate, glutamate, and saccharopine, and decreased cystine provided evidence of reductive stress. Seventy-five percent of the metabolomic abnormalities found were individualized. Personalized deficiencies in CoQ10, flavin adenine dinucleotide (FAD), citrulline, lutein, carnitine, or folate were found. Pathways regulated by mitochondrial function dominated the metabolic signature. Peripheral blood metabolomics identified mitochondrial dysfunction and reductive stress as common denominators in suicidal ideation associated with treatment-refractory major depressive disorder. Individualized metabolic differences were found that may help with personalized management.

The causal association between obesity and gastric cancer and shared molecular signatures: a large-scale Mendelian randomization and multi-omics analysis.

Purpose: While observational studies have identified obesity as a potential risk factor for gastric cancer, the causality remains uncertain. This study aimed to evaluate the causal relationship between obesity and gastric cancer and identify the shared molecular signatures linking obesity to gastric cancer. Methods: A two-sample Mendelian randomization (MR) analysis was conducted using the GWAS data of body fat percentage (exposure, n = 331,117) and gastric cancer (outcome, n = 202,308). Bioinformatics and meta-analysis of multi-omics data were performed to identify key molecules mediating the causality. The meta-analysis of the plasma/serum proteome included 1,662 obese and 3,153 gastric cancer patients. Obesity and gastric cancer-associated genes were identified using seven common gene ontology databases. The transcriptomic data were obtained from TCGA and GEO databases. The Bioinformatic findings were clinically validated in plasma from 220 obese and 400 gastric cancer patients across two hospitals. Finally, structural-based virtual screening (SBVS) was performed to explore the potential FDA-approved drugs targeting the identified mediating molecules. Results: The MR analysis revealed a significant causal association between obesity and gastric cancer (IVW, OR = 1.37, 95% CI:1.12-1.69, P = 0.0028), without pleiotropy or heterogeneity. Bioinformatic and meta-analysis of multi-omics data revealed shared TNF, PI3K-AKT, and cytokine signaling dysregulation, with significant upregulation of AKT1, IL-6, and TNF. The clinical study confirmed widespread upregulation of systemic inflammatory markers in the plasma of both diseases. SBVS identified six novel potent AKT1 inhibitors, including the dietary supplement adenosine, representing a potentially preventive drug with low toxicity. Conclusion: Obesity causally increases gastric cancer, likely mediated by persistent AKT1/IL-6/TNF upregulation. As a potential AKT1 inhibitor, adenosine may mitigate the obesity-to-gastric cancer transition. These findings could inform preventive drug development to reduce gastric cancer risk in obesity.

Dual CNN cross-teaching semi-supervised segmentation network with multi-kernels and global contrastive loss in ACDC.

The cross-teaching based on Convolutional Neural Network (CNN) and Transformer has been successful in semi-supervised learning; however, the information interaction between local and global relations ignores the semantic features of the medium scale, and at the same time, the information in the process of feature coding is not fully utilized. To solve these problems, we proposed a new semi-supervised segmentation network. Based on the principle of complementary modeling information of different kernel convolutions, we design a dual CNN cross-supervised network with different kernel sizes under cross-teaching. We introduce global feature contrastive learning and generate contrast samples with the help of dual CNN architecture to make efficient use of coding features. We conducted plenty of experiments on the Automated Cardiac Diagnosis Challenge (ACDC) dataset to evaluate our approach. Our method achieves an average Dice Similarity Coefficient (DSC) of 87.2% and Hausdorff distance ([Formula: see text]) of 6.1 mm on 10% labeled data, which is significantly improved compared with many current popular models. Supervised learning is performed on the labeled data, and dual CNN cross-teaching supervised learning is performed on the unlabeled data. All data would be mapped by the two CNNs to generate features, which are used for contrastive learning to optimize the parameters.

EFC-YOLO: An Efficient Surface-Defect-Detection Algorithm for Steel Strips.

The pursuit of higher recognition accuracy and speed with smaller model sizes has been a major research topic in the detection of surface defects in steel. In this paper, we propose an improved high-speed and high-precision Efficient Fusion Coordination network (EFC-YOLO) without increasing the model's size. Since modifications to enhance feature extraction in shallow networks tend to affect the speed of model inference, in order to simultaneously ensure the accuracy and speed of detection, we add the improved Fusion-Faster module to the backbone network of YOLOv7. Partial Convolution (PConv) serves as the basic operator of the module, which strengthens the feature-extraction ability of shallow networks while maintaining speed. Additionally, we incorporate the Shortcut Coordinate Attention (SCA) mechanism to better capture the location information dependency, considering both lightweight design and accuracy. The de-weighted Bi-directional Feature Pyramid Network (BiFPN) structure used in the neck part of the network improves the original Path Aggregation Network (PANet)-like structure by adding step branches and reducing computations, achieving better feature fusion. In the experiments conducted on the NEU-DET dataset, the final model achieved an 85.9% mAP and decreased the GFLOPs by 60%, effectively balancing the model's size with the accuracy and speed of detection.