Study on the factors affecting the mechanical properties of antler under quasi-static compression.

In the natural environment, antlers have a significant mechanical structure that protects the deer head from injury. In this paper, the mechanical properties of antlers were evaluated through quasi-static compression tests and microstructural observation of samples from antlers, and the relationship between the sampling position, load direction, and microstructure and their mechanical properties were investigated. Compression experiments confirmed that the tines had the strongest mechanical properties, followed by the main beams and finally the brow tines. The mechanical properties of the test specimens subjected to axial compression were higher than those of lateral compression. The axial load test of the longitudinal sample of the tine of the antler has the best mechanical properties. Its specific energy absorption is 51.33 J/g, the peak crushing force is 1.26 kN, and the mean crushing force is 1.47 kN. There are many tubular structures in the transverse sections of antlers, and the distribution of fibers in the vertical direction is laminar with alternating rows forming a helical structure. Tubular structures were found to be prevalent in some of the better biomechanical structures by comparison. Numerical modeling simulations to describe the effect of tubular structures on the mechanical properties of antlers. The simulation results show that the increase in the size of the tubular structure improves its energy absorption within the variation of a 20% increase in the size of the long and short axis. These findings provide a theoretical and experimental basis for the design of energy-absorbing structures. RESEARCH HIGHLIGHTS: In this paper, transverse and longitudinal samples were taken from the main beam, tine, and brow tine of antlers, and axial and lateral compression were carried out, respectively. In this paper, quasi-static compression experiments were carried out on antler samples to study the effects of sampling position, loading direction, and microstructure on antler mechanical properties. By means of microstructure observation and numerical modeling, it is determined that the size of the antler Havel tube has an effect on its mechanical properties.

Targeting microbiota-immune-synaptic plasticity to explore the effect of tea polyphenols on improving memory in the aged type 2 diabetic rat model.

OBJECTIVES: The study aimed to explore whether TP could improve memory in the aged type 2 diabetic rat model by regulating microbiota-immune-synaptic plasticity axis. METHODS: The experiment was divided into two parts. Firstly, to investigate the effects of TP on the physiopathology of the aged T2DM model rats, rats were randomly divided into the Normal control group, the aged group, the Aged T2DM model group, the TP 75, 150, 300 mg/kg groups, the 150 mg/kg Piracetam group and the 3 mg/kg Rosiglitazone group. Then, to further verify whether TP improved memory in aged T2DM rat model by regulating intestinal flora, the fecal microbiota transplantation (FMT) from the rats in the 300 mg/kg TP group into the rats in the aged T2DM model group was carried out. Effects on gut microbiota, colonic integrity (epithelial tight junction proteins), and endotoxemia (serum LPS) were examined, along with synaptic structure, synaptic plasticity-related structural proteins and inflammation signaling of the hippocampus in our study. RESULTS: Our results demonstrated that TP alleviated memory impairments in the aged T2DM rat model. The specific outcomes were as follows: TP 300 mg/kg corrected the gut dysbacteriosis, alleviated intestinal permeability reduction and peripheral/central inflammation, inhibited the TLR4/NF-κB signaling pathway. Meanwhile, TP improved the synaptic plasticity in the hippocampus of the aged T2DM model rats, whose expressions of SYN, PSD 95, NMDAR1 and GluR1 in hippocampus were significantly up-regulated. Surprisingly, rats of the FMT group displayed the same changes. DISCUSSION: TP improves the memory in aged T2DM rat model. The mechanism may be related to the alteration of gut flora, which can inhibit hippocampal TLR4/NF-κB signaling to attenuate neuroinflammation, then improve synaptic plasticity. The study proposes that TP interventions aimed at manipulating the gut microbiota may hold great potential as an effective approach for preventing and treating this disease.

[Tea polyphenols improves depression-like behavior in aged type 2diabetes rats by regulating microglia polarization].

OBJECTIVE: To investigate the effect of tea polyphenols(TP) on improving depression-like behavior in aged type 2 diabetes(T2DM) model rats. METHODS: A total of 40 8-week-old SD male rats were randomly divided into the control group(n=10) and the modeling group(n=30) according to the body weight. The rats in the modeling group were fed with high-glucose and high-fat diet and treated with 50 mg/kg D-galactose by intraperitoneal injection daily until the end of the experiment, while the rats in the control group were fed with the standard diet and treated with an equal volume of saline by intraperitoneal injection. After 4 weeks, the rats in the modeling group were injected with 25 mg/kg STZ, meanwhile the rats in the control group were injected with an equal volume of citric acid buffer. The level of fasting blood glucose(FBG) was measured on the 14~(th) day. When FBG≥16.7 mmol/L, the rats were identified as successful model of the T2DM rats. Then, the model rats were randomly divided into the model group, 150, 300 mg/kg TP groups(n=10, respectively), and the rats were given intragastric intervention for 8 weeks. The levels of the FBG were detected, and the depression-like behavior of rats was assessed by the open field test(OFT) and forced swimming test(FST). The density of microglia in hippocampus CA1 region was assessed by immunofluorescence staining, and protein expressions of P53, Iba1, iNOS, Arg-1 and BDNF were determined by western blot. RESULTS: Compared with the control group, the levels of FBG in the rats of the model group were obviously increased(P<0.01). In the OFT, the frequencies of rearing and grooming in the rats of model group markedly was decreased, while in the FST, the immobility time extensively was increased(P<0.01). The density of microglia in hippocampus CA1 region was increased(P<0.01). The expressions of P53, Iba1 and iNOS were increased, and the expressions of Arg-1 and BDNF were decreased(P<0.01). Additionally, compared with the model group, in the OFT, the frequencies of rearing and grooming were increased in the rats in 150 and 300 mg/kg TP group(P<0.01). The density of microglia in hippocampus CA1 region was decreased(P<0.01). The expressions of P53, Iba1 and iNOS were down-regulated, and the expression of BDNF was up-regulated(P<0.01). Additionally, compared with the model group, the levels of FBG was decreased in the rats in the 300 mg/kg TP group(P<0.01). The immobility time was decreased in the FST(P<0.01). The expression of Arg-1 was down-regulated(P<0.01). CONCLUSION: TP can improve depression-like behavior in aged T2DM model rats, and its mechanism may be related to regulate microglia M1/M2 polarization and up-regulate expression of BDNF in hippocampus.

UCL-Dehaze: Toward Real-World Image Dehazing via Unsupervised Contrastive Learning.

While the wisdom of training an image dehazing model on synthetic hazy data can alleviate the difficulty of collecting real-world hazy/clean image pairs, it brings the well-known domain shift problem. From a different yet new perspective, this paper explores contrastive learning with an adversarial training effort to leverage unpaired real-world hazy and clean images, thus alleviating the domain shift problem and enhancing the network's generalization ability in real-world scenarios. We propose an effective unsupervised contrastive learning paradigm for image dehazing, dubbed UCL-Dehaze. Unpaired real-world clean and hazy images are easily captured, and will serve as the important positive and negative samples respectively when training our UCL-Dehaze network. To train the network more effectively, we formulate a new self-contrastive perceptual loss function, which encourages the restored images to approach the positive samples and keep away from the negative samples in the embedding space. Besides the overall network architecture of UCL-Dehaze, adversarial training is utilized to align the distributions between the positive samples and the dehazed images. Compared with recent image dehazing works, UCL-Dehaze does not require paired data during training and utilizes unpaired positive/negative data to better enhance the dehazing performance. We conduct comprehensive experiments to evaluate our UCL-Dehaze and demonstrate its superiority over the state-of-the-arts, even only 1,800 unpaired real-world images are used to train our network. Source code is publicly available at https://github.com/yz-wang/UCL-Dehaze.

Technology-enhanced higher education: Text mining and bibliometrics.

Objectives: Research on technology-enhanced higher education (TEHE) has been active and influential in educational technology. The study had three objectives: (1) to recognize the tendencies in the field and the contributing countries/regions/institutions, (2) to visualize scientific collaborations, and (3) to reveal important research topics, their developmental tendencies, correlations, and distributions across contributing countries/regions/institutions. Methods: We collected 609 papers in relation to TEHE from 2004 to 2022 and analyzed them using text mining and bibliometric methods. Specifically, we focused on determining article trends, identifying contributing institutions/countries/regions, visualizing scientific collaborations through social network analysis, and revealing the important topics and their conceptual evolutions over time using topic models, Mann-Kendall trend test, hierarchical clustering, and Sankey visualization. Results: Regarding the first objective, TEHE articles have grown consistently and will continue to expand. This growth was due to the contributions of Spanish universities and institutions from other countries/regions such as the USA, the UK, Australia, Germany, China, and Turkey. Regarding the second objective, the exploration of regional and institutional collaborations through social networks revealed that geographically adjacent institutions tended to foster close collaborations, particularly among those sharing similar research interests. Nevertheless, more cross-regional collaborations are needed to advance TEHE research. Regarding the third objective, the analysis of topics highlighted research hotspots and emerging themes such as Massive Online Open Courses, AI and big data in education, Gamification and engagement, Learning effectiveness and strategies, Social networks and discussion forums, COVID-19 and online learning, and Plagiarism detection and learning analytics. Conclusions: This bibliometric study comprehensively analyzed the research landscape of TEHE research regarding contributors, collaborations, and research topics, and offers a glimpse into what the future may hold. It can be used as a guide for contributors to the field to identify the current research hotspots and emerging themes.

Targeting ERS-mitophagy in hippocampal neurons to explore the improvement of memory by tea polyphenols in aged type 2 diabetic rats.

Increasing evidence demonstrated that mitophagy and endoplasmic reticulum stress (ERS) was closely associated with memory decline in elderly type 2 diabetes mellitus (T2DM). Tea polyphenols (TP), an excellent natural antioxidant, has been reported to have neuroprotective properties in aging and diabetes, but the underlying mechanisms are still not fully understood. This study targets ERS-mitophagy in hippocampal neurons to investigate the improvement effect of memory in aged T2DM rats by TP. Rats were randomly divided into the control group, the aged group, the aged T2DM model group, the TP 75, 150, 300 mg/kg groups. TP 300 mg/kg ameliorated mitophagy by decreasing the levels of p-mTOR (S2448), P62 and HSP60 and increasing the levels of PINK1 and Parkin, the ratio of LC3Ⅱ/LC3Ⅰ, co-localization of LC3 and HSP60 and the number of autophagosomes and autolysosomes. TP 300 mg/kg attenuated ERS by downregulating the levels of p-PERK, p-eIF2α, ATF4, GRP78 and restoring the ER structure. To further verify epigallocatechin gallate (EGCG), which is the main active component of TP, enhanced mitophagy by inhibiting ERS, PC12 cells were pretreated with ERS activator tunicamycin (TM) or ERS inhibitor 4-phenylbutyric acid (4-PBA). The results showed that the improvement of mitophagy by EGCG was inhibited by TM and promoted by 4-PBA. Collectively, ERS-mitophagy in hippocampal neurons plays a key role in the improvement of memory by TP in aged T2DM rats. This study will provide a new perspective and strategy for the prevention of memory decline in elderly with T2DM.

Mimosa Kinetic Façade: Bio-Inspired Ventilation Leveraging the Mimosa Pudica Mechanism for Enhanced Indoor Air Quality.

In light of pressing global health concerns, the significance of indoor air quality in densely populated structures has been emphasized. This research introduces the Mimosa kinetic façade, an innovative design inspired by the adaptive responsiveness of the Mimosa plant to environmental stimuli. Traditional static architectural façades often hinder natural ventilation, leading to diminished air quality with potential health and cognitive repercussions. The Mimosa kinetic façade addresses these challenges by enhancing effective airflow and facilitating the removal of airborne contaminants. This study evaluates the façade's impact on quality of life and its aesthetic contribution to architectural beauty, utilizing the biomimicry design spiral for a nature-inspired approach. Computational simulations and physical tests were conducted to assess the ventilation capacities of various façade systems, with a particular focus on settings in Bangkok, Thailand. The study revealed that kinetic façades, especially certain patterns, provided superior ventilation compared to static ones. Some patterns prioritized ventilation, while others optimized human comfort during extended stays. Notably, the most effective patterns of the kinetic façade inspired by the Mimosa demonstrated a high air velocity reaching up to 12 m/s, in contrast to the peak of 2.50 m/s in single-sided façades (traditional façades). This highlights the kinetic façade's potential to rapidly expel airborne particles from indoor spaces, outperforming traditional façades. The findings underscore the potential of specific kinetic façade patterns in enhancing indoor air quality and human comfort, indicating a promising future for kinetic façades in architectural design. This study aims to achieve an optimal balance between indoor air quality and human comfort, although challenges remain in perfecting this equilibrium.

PointSee: Image Enhances Point Cloud.

There is a prevailing trend towards fusing multi-modal information for 3D object detection (3OD). However, challenges related to computational efficiency, plug-and-play capabilities, and accurate feature alignment have not been adequately addressed in the design of multi-modal fusion networks. In this paper, we present PointSee, a lightweight, flexible, and effective multi-modal fusion solution to facilitate various 3OD networks by semantic feature enhancement of point clouds (e.g., LiDAR or RGB-D data) assembled with scene images. Beyond the existing wisdom of 3OD, PointSee consists of a hidden module (HM) and a seen module (SM): HM decorates point clouds using 2D image information in an offline fusion manner, leading to minimal or even no adaptations of existing 3OD networks; SM further enriches the point clouds by acquiring point-wise representative semantic features, leading to enhanced performance of existing 3OD networks. Besides the new architecture of PointSee, we propose a simple yet efficient training strategy, to ease the potential inaccurate regressions of 2D object detection networks. Extensive experiments on the popular outdoor/indoor benchmarks show quantitative and qualitative improvements of our PointSee over thirty-five state-of-the-art methods.

Identification of a potential prognostic model combining pyroptosis-related gene with immune microenvironment for pancreatic ductal adenocarcinoma.

BACKGROUND: Pancreatic ductal adenocarcinoma (PDAC) is a fatal tumor with grave prognosis. Pyroptosis, a programmed cell death, is involved in tumorigenesis. However, a few studies have elucidated the functions of pyroptosis in PDAC. METHODS: The mRNA expression profiles were downloaded from the TCGA and GEO databases. Univariate and LASSO Cox regression analyses were used to screen out differentially expressed genes (DEGs) and construct the pyroptosis-related genes (PRGs) risk model. The efficiency of model was examined by Kaplan-Meier curve, ROC curve, and nomogram. Univariate and multivariate Cox regression analyses were utilized to assess whether the risk model could be used as an independent prognostic factor. The biological function was analyzed by GO, KEGG, and GSEA enrichment analysis. qRT-PCR and immunohistochemical staining detected gene expression. RESULTS: Totally 9 PRGs with differential expression were identified between normal and PDAC tissues. Then, according to PRGs, we filtered out three key DEGs and constructed the prognostic risk model. Kaplan-Meier curve, ROC curve, and nomogram indicated that the prognostic risk model had high survival prediction efficiency. Meanwhile, the risk model had also shown to be an independent prognostic factor. Further functional enrichment analysis showed that cell adhesion, PI3K-AKT signaling pathway, and dysregulated immune status may be associated with PDAC development. External validation of the model was carried out in the GEO cohort, and the results were similar to that in the TCGA cohort. Finally, the expression of three genes was verified by qRT-PCR and immunohistochemical staining. CONCLUSION: The prognostic risk model established in this study can give a good prediction of the prognosis of PDAC patients, which might provide insights into clinical treatments and prognostic prediction of PDAC.

Trust-aware conditional adversarial domain adaptation with feature norm alignment.

Adversarial learning has proven to be an effective method for capturing transferable features for unsupervised domain adaptation. However, some existing conditional adversarial domain adaptation methods assign equal importance to different samples, ignoring the fact that hard-to-transfer samples might damage the conditional adversarial adaptation procedure. Meanwhile, some methods can only roughly align marginal distributions across domains, but cannot ensure category distributions alignment, causing classifiers to make uncertain or even wrong predictions for some target data. Furthermore, we find that the feature norms of real images usually follow a complex distribution, so directly matching the mean feature norms of two domains cannot effectively reduce the statistical discrepancy of feature norms and may potentially induce feature degradation. In this paper, we develop a Trust-aware Conditional Adversarial Domain Adaptation (TCADA) method for solving the aforementioned issues. To quantify data transferability, we suggest utilizing posterior probability modeled by a Gaussian-uniform mixture, which effectively facilitates conditional domain alignment. Based on this posterior probability, a confidence-guided alignment strategy is presented to promote precise alignment of category distributions and accelerate the learning of shared features. Moreover, a novel optimal transport-based strategy is introduced to align the feature norms and facilitate shared features becoming more informative. To encourage classifiers to make more accurate predictions for target data, we also design a mixed information-guided entropy regularization term to promote deep features being away from the decision boundaries. Extensive experiments show that our method greatly improves transfer performance on various tasks.

Federated Domain Adaptation via Transformer for Multi-Site Alzheimer's Disease Diagnosis.

In multi-site studies of Alzheimer's disease (AD), the difference of data in multi-site datasets leads to the degraded performance of models in the target sites. The traditional domain adaptation method requires sharing data from both source and target domains, which will lead to data privacy issue. To solve it, federated learning is adopted as it can allow models to be trained with multi-site data in a privacy-protected manner. In this paper, we propose a multi-site federated domain adaptation framework via Transformer (FedDAvT), which not only protects data privacy, but also eliminates data heterogeneity. The Transformer network is used as the backbone network to extract the correlation between the multi-template region of interest features, which can capture the brain abundant information. The self-attention maps in the source and target domains are aligned by applying mean squared error for subdomain adaptation. Finally, we evaluate our method on the multi-site databases based on three AD datasets. The experimental results show that the proposed FedDAvT is quite effective, achieving accuracy rates of 88.75%, 69.51%, and 69.88% on the AD vs. NC, MCI vs. NC, and AD vs. MCI two-way classification tasks, respectively.

Effects of different timing selections of labor analgesia for primiparae on parturition and neonates.

OBJECTIVE: To explore the effects of labor analgesia for primiparae with different stages of cervical dilation on parturition and neonates. MATERIALS AND METHODS: In the past three years, 530 cases of primiparae who had delivered in the Second People's Hospital of Hefei and were eligible for a vaginal trial of parturition were enrolled as the research subjects. Of these, 360 puerperae had labor analgesia, and the remaining 170 were taken as the control group. Those given labor analgesia were divided into three groups based on the different stages of cervical dilation at that time. There were 160 cases in Group I (cervical dilation <3 cm), 100 cases in Group II (cervical dilation of 3-4 cm), and 100 cases in Group III (cervical dilation of 4-6 cm). The labor and neonatal outcomes were compared among the four groups. RESULTS: The first, second, and total stages of labor in the three groups receiving labor analgesia were all longer than in the control group, and the differences were statistically significant (p < 0.05 in all). Group I had the longest duration of each stage and the total stage of labor. The differences in labor stages and the total stage of labor were not statistically significant between Group II and Group III (P > 0.05). In the three groups with labor analgesia, the usage rate of oxytocin was higher than in the control group, and the differences were statistically significant (P < 0.05). The differences in the incidence of postpartum hemorrhage, the incidence of postpartum urine retention, and the episiotomy rate were not statistically significant among the four groups (P > 0.05). The differences in the neonatal Apgar score were not statistically significant among the four groups (P > 0.05). CONCLUSION: Labor analgesia might prolong the stages of labor but does not affect the neonatal outcomes. It would be optimal to conduct labor analgesia when cervical dilation reaches 3-4 cm.

CSDN: Cross-Modal Shape-Transfer Dual-Refinement Network for Point Cloud Completion.

How will you repair a physical object with some missings? You may imagine its original shape from previously captured images, recover its overall (global) but coarse shape first, and then refine its local details. We are motivated to imitate the physical repair procedure to address point cloud completion. To this end, we propose a cross-modal shape-transfer dual-refinement network (termed CSDN), a coarse-to-fine paradigm with images of full-cycle participation, for quality point cloud completion. CSDN mainly consists of "shape fusion" and "dual-refinement" modules to tackle the cross-modal challenge. The first module transfers the intrinsic shape characteristics from single images to guide the geometry generation of the missing regions of point clouds, in which we propose IPAdaIN to embed the global features of both the image and the partial point cloud into completion. The second module refines the coarse output by adjusting the positions of the generated points, where the local refinement unit exploits the geometric relation between the novel and the input points by graph convolution, and the global constraint unit utilizes the input image to fine-tune the generated offset. Different from most existing approaches, CSDN not only explores the complementary information from images but also effectively exploits cross-modal data in the whole coarse-to-fine completion procedure. Experimental results indicate that CSDN performs favorably against twelve competitors on the cross-modal benchmark.

Omnidirectional Haptic Stimulation System via Pneumatic Actuators for Presence Presentation.

Recently, remote meetings and work-from-home have become more common, reducing the opportunities for face-to-face communication. To facilitate communication among remote workers, researchers have focused on virtual space technology and spatial augmented reality technology. Although these technologies can enhance immersiveness in collaborative work, they face the challenge of fostering a sense of physical contact. In this work, we aimed to foster a sense of presence through haptic stimulation using pneumatic actuators. Specifically, we developed a choker-type wearable device that presents various pressure patterns around the neck; the pattern presented depends on the message the device must convey. Various combinations of haptic presentation are achieved by pumping air to the multiple pneumatic actuators attached to the choker. In addition, we conducted experiments involving actuators of different shapes to optimize the haptic presentation. When linked with a smartphone, the proposed device can present pressure patterns to indicate incoming calls and notifications, to give warning about an obstacle that one who is texting might miss while walking, and to provide direction to a pedestrian. Furthermore, the device can be used in a wide range of applications, from those necessary in daily living to those that enhance one's experience in the realm of entertainment. For example, haptic feedback that synchronizes with the presence of a singer or with the rhythm of a song one listens to or with a performer's movements during a stage performance will immerse users in an enjoyable experience.

Transpapillary Stenting Improves Treatment Outcomes in Patients Undergoing Endoscopic Transmural Drainage of Ductal Disruption-Associated Pancreatic Fluid Collections.

INTRODUCTION: Endoscopic transmural drainage (TMD) has been accepted as the preferred therapy for symptomatic pancreatic fluid collections (PFCs). Recurrence of PFCs presents a unique challenge in patients with disrupted pancreatic duct (PD). We aimed to evaluate whether transpapillary drainage (TPD) provides additional benefits to TMD in patients with PD disruption. METHODS: This was a multicenter retrospective study. Consecutive patients who underwent TMD, TPD, or combined drainage (CD) of PFCs were included. The primary outcome was to compare PFC recurrence among different groups. The secondary outcomes were the technical success rate, length of hospital stay, and procedure-related complications. RESULTS: A total of 153 patients, which consists of 57 patients with pancreatic pseudocysts and 96 patients with walled-off necrosis, were included. PFC recurrence was more common in patients with PD disruption than those with an intact main duct (19% vs 1.4%, P < 0.001). PD disruption was identified as a major risk factor of PFC recurrence by univariable and multivariable analyses. The recurrence rate of CD was significantly lower than TMD only or TPD only (6.5% vs 15.4% vs 22.7%, P < 0.01). The length of hospital stay of CD was significantly shorter than TMD only or TPD only (5 [3.0-9.0] vs 7.0 [5.0-12.0] vs 9 [7.0-16.0], P < 0.001). Dual-modality drainage did not increase procedure-related complications compared with TMD only (13.0% vs 12.8%, P > 0.05). Partial PD disruption was bridged in 87.3% cases while complete PD disruption was reconnected in 55.2% cases. Although statistically not significant, the clinical success rate in walled-off necrosis cases with actively bridged ducts was much higher than those with passively bridged ducts (76.9% vs 40%). DISCUSSION: Transpapillary pancreatic duct stenting seems to improve the efficacy of endoscopic TMD of pancreatic duct disruption-associated PFCs by reducing the recurrence rate and shortening the length of hospital stay.

Preoperative Extrapancreatic Extension Prediction in Patients with Pancreatic Cancer Using Multiparameter MRI and Machine Learning-Based Radiomics Model.

RATIONALE AND OBJECTIVES: Pancreatic cancer is a common malignant tumor with a dismal prognosis. Preoperative differentiation of extrapancreatic extension (EPE) based on radiomics will facilitate treatment decision-making. MATERIALS AND METHODS: This research retrospectively recruited 156 patients from two medical centers. 122 patients from the center A were randomly divided into the training set and the internal test set in a 4:1 ratio. Additionally, 34 patients from the center B served as the external test set. Radiomics features were extracted from multiparametric MRI (MP-MRI), containing axial T2 weighted imaging (T2WI), diffusion weighted imaging (DWI), and dynamic contrast enhancement (DCE) sequences. The three-step method was used for feature extraction: SelecteKBest, least absolute shrinkage and selection operator (LASSO) algorithm, and recursive feature elimination based on random forest (RFE-RF). The model was constructed using six classifiers based on machine learning, and the classifier with the best performance was chosen. Finally, clinical factors associated with EPE were incorporated into the combined model. RESULTS: The classifier with the best performance was XGBoost, which obtained area under curve (AUC) values of 0.853 and 0.848 in the internal and external test sets, respectively. Through SelectKBest, the most relevant clinical factor for EPE was determined to be platelet, which was then added to the combined model, yielding AUC values of 0.880 and 0.848 in the internal and external test sets, respectively. CONCLUSION: Radiomics models had the potential to noninvasively and accurately predict EPE before surgery. Additionally, it would add value to personalized precision treatment.

GeoDualCNN: Geometry-Supporting Dual Convolutional Neural Network for Noisy Point Clouds.

We propose a geometry-supporting dual convolutional neural network (GeoDualCNN) for both point cloud normal estimation and denoising. GeoDualCNN fuses the geometry domain knowledge that the underlying surface of a noisy point cloud is piecewisely smooth with the fact that a point normal is properly defined only when local surface smoothness is guaranteed. Centered around this insight, we define the homogeneous neighborhood (HoNe) which stays clear of surface discontinuities, and associate each HoNe with a point whose geometry and normal orientation is mostly consistent with that of HoNe. Thus, we not only obtain initial estimates of the point normals by performing PCA on HoNes, but also for the first time optimize these initial point normals by learning the mapping from two proposed geometric descriptors to the ground-truth point normals. GeoDualCNN consists of two parallel branches that remove noise using the first geometric descriptor (a homogeneous height map, which encodes the point-position information), while preserving surface features using the second geometric descriptor (a homogeneous normal map, which encodes the point-normal information). Such geometry-supporting network architectures enable our model to leverage previous geometry expertise and to benefit from training data. Experiments with noisy point clouds show that GeoDualCNN outperforms the state-of-the-art methods in terms of both noise-robustness and feature preservation.

Refine-Net: Normal Refinement Neural Network for Noisy Point Clouds.

Point normal, as an intrinsic geometric property of 3D objects, not only serves conventional geometric tasks such as surface consolidation and reconstruction, but also facilitates cutting-edge learning-based techniques for shape analysis and generation. In this paper, we propose a normal refinement network, called Refine-Net, to predict accurate normals for noisy point clouds. Traditional normal estimation wisdom heavily depends on priors such as surface shapes or noise distributions, while learning-based solutions settle for single types of hand-crafted features. Differently, our network is designed to refine the initial normal of each point by extracting additional information from multiple feature representations. To this end, several feature modules are developed and incorporated into Refine-Net by a novel connection module. Besides the overall network architecture of Refine-Net, we propose a new multi-scale fitting patch selection scheme for the initial normal estimation, by absorbing geometry domain knowledge. Also, Refine-Net is a generic normal estimation framework: 1) point normals obtained from other methods can be further refined, and 2) any feature module related to the surface geometric structures can be potentially integrated into the framework. Qualitative and quantitative evaluations demonstrate the clear superiority of Refine-Net over the state-of-the-arts on both synthetic and real-scanned datasets.

The Wet Suction Technique Enhances the Diagnostic Efficacy and Aspirate Quality of EUS-FNA for Solid Lesions: A Multicenter Retrospective Study in China.

GOALS: To comprehensively compare the wet suction technique with the conventional dry suction technique for endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) in solid lesions. BACKGROUND: Optimal suction techniques for EUS-FNA remain uncertain when approaching solid lesions. STUDY: We performed a retrospective study of EUS-FNA at 3 medical centers in China. A total of 203 patients were enrolled who received 2 passes of EUS-FNA with 22-G needles. If the first pass underwent dry suction, the second pass was wet suction. Otherwise, the order of suction technique is opposite. Diagnostic accuracy, sample quality (including cellularity and blood contamination), and sample quantity (including specimen adequacy, the maximum intact specimen length, and the total specimen length) were compared between wet-suction and dry-suction techniques. RESULTS: The patients included 143 pancreatic lesions and 60 nonpancreatic lesions. Compared with the dry suction technique, the wet suction technique yielded a significantly higher diagnostic accuracy (85.22% vs. 72.41%, P =0.002), better specimen adequacy score and cellularity score ( P <0.0001), and lower blood contamination score ( P <0.0001). In the subgroup analysis, wet suction provided significantly higher diagnostic accuracy in pancreatic cancer without chronic pancreatitis ( P <0.05), and better cellularity score and specimen adequacy score, lower blood contamination score, and longer maximum intact specimen length and total specimen length in various lesions than that in dry suction. CONCLUSIONS: The wet suction technique resulted in significantly higher diagnostic accuracy in pancreatic cancer without chronic pancreatitis, and better cellularity and histologic specimen in most of solid lesions.