A Nomogram Including Total Cerebral Small Vessel Disease Burden Score for Predicting Mild Vascular Cognitive Impairment in Patients with Type 2 Diabetes Mellitus.

Background: Total cerebral small vessel disease (CSVD) burden score is an important predictor of vascular cognitive impairment (VCI). However, few predictive models of VCI in type 2 diabetes mellitus (T2DM) patients have included the total CSVD burden score, especially in the early stage of VCI. Objective: To develop and validate a nomogram that includes the total CSVD burden score to predict mild VCI in patients with T2DM. Methods: A total of 322 eligible participants with T2DM who were divided into mild and normal cognitive groups were enrolled in this retrospective study. Demographic data, laboratory data and imaging markers of CSVD were collected. The total CSVD burden score was calculated by combining the different CSVD markers. Step-backward multivariable logistic regression analysis with the Akaike information criterion was applied to select significant predictors and develop a best-fit predictive nomogram. The performance of the nomogram was assessed in terms of discriminative ability, calibrated ability, and clinical usefulness. Results: The nomogram model consisted of five variables: age, education, hemoglobin A1c level, serum homocysteine level, and total CSVD burden score. A nomogram with these variables showed good discriminative ability (area under the receiver operating characteristic curve was 0.801 in internal verification). In addition, the Hosmer-Lemeshow test (χ2 =9.226, P=0.417) and bootstrap-corrected calibration plot indicated that the nomogram had good calibration. The Brier score of the predictive model was 0.178. Decision curve analysis demonstrated that when the threshold probability ranged between 16% and 98%, the use of the nomogram to predict mild VCI in patients with T2DM provide a greater net benefit. Conclusions: The nomogram, composed of age, education, stroke, HbA1c level, Hcy level, and total CSVD burden score, had good predictive accuracy and may provide clinicians with a practical tool for predicting the risk of mild VCI in T2DM patients.

Nonlinear relationship between triglycerides and cognitive function after acute ischemic stroke among older adults.

Background: Although studies have explored the association between triglyceride levels and cognitive function after acute ischemic stroke (AIS), the results have been conflicting. Therefore, the purpose of this study was to investigate the relationship between triglyceride levels and cognitive function after AIS among older adults. Methods: This is an observational cross-sectional study. From November 2022 to June 2023, we consecutively collected patients diagnosed with AIS in China. Triglyceride levels were measured within 24 h of admission. The Mini-Mental State Examination (MMSE) was used to assess cognitive function. Nonlinear associations between triglyceride levels and cognitive function were assessed using smooth curve fitting and threshold effect analysis. Results: In this study, a total of 221 patients (mean ± SD: 70.64 ± 7.43 years) with AIS were consecutively recruited, among whom 144 (65.16%) were male. Among the 221 recruited patients, 102 (46.15%) had cognitive impairment. Triglyceride levels and cognitive impairment were found to have a nonlinear association after controlling for potential confounders, with an inflection point at 0.8 mmol/L. Below the inflection point, triglyceride levels were positively correlated with MMSE scores (ß = 14.11, 95% confidence interval [CI] = 2.33-25.89, P = 0.020). However, above the inflection point, the correlation between MMSE score and triglyceride levels was not statistically significant (ß = 1.04, 95% CI = -1.27 - 3.34, P = 0.380). Conclusion: There is a nonlinear association between triglyceride levels and cognitive function after AIS in older adults. Triglyceride was positively connected with cognitive function when it was less than 0.8 mmol/L.

Generalizable and Discriminative Representations for Adversarially Robust Few-Shot Learning.

Few-shot image classification (FSIC) is beneficial for a variety of real-world scenarios, aiming to construct a recognition system with limited training data. In this article, we extend the original FSIC task by incorporating defense against malicious adversarial examples. This can be an arduous challenge because numerous deep learning-based approaches remain susceptible to adversarial examples, even when trained with ample amounts of data. Previous studies on this problem have predominantly concentrated on the meta-learning framework, which involves sampling numerous few-shot tasks during the training stage. In contrast, we propose a straightforward but effective baseline via learning robust and discriminative representations without tedious meta-task sampling, which can further be generalized to unforeseen adversarial FSIC tasks. Specifically, we introduce an adversarial-aware (AA) mechanism that exploits feature-level distinctions between the legitimate and the adversarial domains to provide supplementary supervision. Moreover, we design a novel adversarial reweighting training strategy to ameliorate the imbalance among adversarial examples. To further enhance the adversarial robustness without compromising discriminative features, we propose the cyclic feature purifier during the postprocessing projection, which can reduce the interference of unforeseen adversarial examples. Furthermore, our method can obtain robust feature embeddings that maintain superior transferability, even when facing cross-domain adversarial examples. Extensive experiments and systematic analyses demonstrate that our method achieves state-of-the-art robustness as well as natural performance among adversarially robust FSIC algorithms on three standard benchmarks by a substantial margin.

Knocking out FAM20C in pre-osteoblasts leads to up-regulation of osteoclast differentiation to affect long bone development.

Family with sequence similarity 20 member C (FAM20C) is a Golgi casein kinase that phosphorylates extracellularly-secreted regulatory proteins involved in bone development and mineralization, but its specific role in bone development is still largely unknown. In this study, to examine the specific mechanisms that FAM20C influences bone development, we cross-bred Osx-Cre with FAM20Cflox/flox mice to establish a Osx-Cre; FAM20Cflox/flox knockout (oKO) mouse model; FAM20C was KO in pre-osteoblasts. oKO development was examined at 1-10 weeks, in which compared to control FAM20Cflox/flox, they had lower body weights and bone tissue mineralization. Furthermore, oKO had lower bone volume fractions, thickness, and trabecular numbers, along with higher degrees of trabecular separation. These mice also had decreased femoral metaphyseal cartilage proliferation layer, along with thickened hypertrophic layer and increased apoptotic cell counts. Transcriptomic analysis found that differentially-expressed genes in oKO were concentrated in the osteoclast differentiation pathway, in line with increased osteoclast presence. Additionally, up-regulation of osteoclast-related, and down-regulation of osteogenesis-related genes, were identified, in which the most up-regulated genes were signal regulatory protein ß-1 family (Sirpb1a-c) and mitogen-activated protein kinase 13. Overall, FAM20C KO in pre-osteoblasts leads to abnormal long bone development, likely due to subsequent up-regulation of osteoclast differentiation-associated genes.

Increasing the sensor channels: a solution for the pressing offsets that cause the physiological parameter inaccuracy in radial artery pulse signal acquisition.

Introduction: In studies of pulse wave analysis, single-channel sensors only adopt single temporal pulse signals without spatial information to show pulse-feeling patterns. Multi-channel arterial pulse signals, also named as three-dimensional pulse images (3DPIs), provide the spatial and temporal characteristics of radial pulse signals. When involving single or few-channel sensors, pressing offsets have substantial impacts on obtaining inaccurate physiological parameters like tidal peak (P2). Methods: This study discovers the pressing offsets in multi-channel pulse signals and analyzes the relationship between the pressing offsets and time of P2 (T2) by qualifying the pressing offsets. First, we employ a data acquisition system to capture 3DPIs. Subsequently, the errorT2 is developed to qualify the pressing offsets. Results: The outcomes display a central low and peripheral high pattern. Additionally, the errorT2 increase as the distances from the artery increase, particularly at the radial ends of the blood flow direction. For every 1 mm increase in distances between sensing elements and center sensing elements, the errorT2 in the radial direction escalates by 4.87%. When the distance is greater than 3.42 mm, the errorT2 experiences a sudden increase. Discussion: The results show that increasing the sensor channels can overcome the pressing offsets in radial pulse signal acquisition.

A reliable evaluation approach for multichannel signal denoising algorithms based on a novel arterial pulse acquisition system.

Background: Tactile sensors are utilized to measure multichannel pulse signals in pulse wave analysis (PWA). Owing to noise interferences, researchers have applied various denoising algorithms on multichannel pulse signals. To comprehensively assess these algorithms, numerous evaluation metrics have been proposed. However, these studies did not investigate the noise mechanisms in depth and lacked reference pulse signals, thus making the evaluations insufficiently objective. Materials and methods: An applicable denoising evaluation approach for multichannel pulse signal algorithms based on an arterial pulse acquisition system is established by superimposing real-world multichannel noise to the reference signals. The system, comprising a SphygmoCor and a uniaxial noise acquisition device, allows us to acquire single-reference pulse signals as well as real-world multichannel noise. Results: We assess eight popular denoising algorithms with three evaluation metrics, including amplitude relative error (ARE), mean square error (MSE) and increased percentage signal-noise ratio (SNR%). Our proposed approach provides accurate and objective evaluations of multichannel pulse signal denoising. Notably, classic algorithms for single-channel denoising are not recommended for multichannel denoising. Comparatively, RPCA-based algorithms can denoise pulse signals independently for each channel. Conclusion: This study sets the stage for the establishment of accurate and objective pulse signal denoising evaluations and provides insights for data-driven clinical diagnoses in cardiovascular medicine.

Growth of infants delivered by mothers with HIV in Guangxi, China: An 18-month longitudinal follow-up study, 2015-2021.

OBJECTIVES: The prevention of mother-to-child transmission of HIV has been a global success. But little is known about the growth parameters of infants delivered by mothers with HIV or the drug resistance of infants with HIV in China. The study aimed to assess growth parameters and drug resistance in Chinese infants exposed to HIV. METHODS: We conducted an 18-month longitudinal follow-up study of 3283 infants (3222 without HIV; 61 with HIV) born to mothers with HIV in the Guangxi Zhuang Autonomous Region between January 2015 and December 2021. The weight and length of all participants was recorded. In addition, genetic subtypes and drug resistance analysis were performed for infants with HIV. RESULTS: Compared with infants without HIV, those with HIV had significantly lower weight/length Z-scores, except at 18 months of age. The length/age Z-scores of infants with HIV was significantly reduced, except at 1 month of age. The weight/age Z-scores of infants with HIV were significantly lower at all follow-up time points. The weight/length Z-scores of male infants without HIV were significantly lower than for female infants without HIV at all follow-up time points. Male infants without HIV had lower length/age and weight/age Z-scores than female infants at the remaining follow-up points, except at 1 month of age. Of a total of 61 infants with HIV, subtype and drug-resistance data were obtained from 37 (60.66%) samples. Infants with HIV were dominated by the CRF01_AE genotype and showed a diversity of mutation sites dominated by non-nucleoside reverse transcriptase inhibitor resistance. CONCLUSION: Our study demonstrates the growth of infants exposed to HIV in southwest China and provides detailed information on subtype distribution and drug resistance of those with HIV. Nutritional support and drug-resistance surveillance for infants exposed to HIV need to be strengthened.

Separable Spatial-Temporal Residual Graph for Cloth-Changing Group Re-Identification.

Group re-identification (GReID) aims to correctly associate group images belonging to the same group identity, which is a crucial task for video surveillance. Existing methods only model the member feature representations inside each image (regarded as spatial members), which leads to potential failures in long-term video surveillance due to cloth-changing behaviors. Therefore, we focus on a new task called cloth-changing group re-identification (CCGReID), which needs to consider group relationship modeling in GReID and robust group representation against cloth-changing members. In this paper, we propose the separable spatial-temporal residual graph (SSRG) for CCGReID. Unlike existing GReID methods, SSRG considers both spatial members inside each group image and temporal members among multiple group images with the same identity. Specifically, SSRG constructs full graphs for each group identity within the batched data, which will be completely and non-redundantly separated into the spatial member graph (SMG) and temporal member graph (TMG). SMG aims to extract group features from spatial members, and TMG improves the robustness of the cloth-changing members by feature propagation. The separability enables SSRG to be available in the inference rather than only assisting supervised training. The residual guarantees efficient SSRG learning for SMG and TMG. To expedite research in CCGReID, we develop two datasets, including GroupPRCC and GroupVC, based on the existing CCReID datasets. The experimental results show that SSRG achieves state-of-the-art performance, including the best accuracy and low degradation (only 2.15% on GroupVC). Moreover, SSRG can be well generalized to the GReID task. As a weakly supervised method, SSRG surpasses the performance of some supervised methods and even approaches the best performance on the CSG dataset.

Unveiling the potent effect of vitamin D: harnessing Nrf2/HO-1 signaling pathways as molecular targets to alleviate urban particulate matter-induced asthma inflammation.

BACKGROUND: Asthma is the most common allergic disease characterized by an inflammatory response in the airways. Mechanismly, urban particulate matter (PM) is the most widely air pollutant associated with increased asthma morbidity and airway inflammation. Current research found that vitamin D is an essential vitamin with anti-inflammatory, antioxidant and other medical efficacy. Inadequate or deficient vitamin D often leads to the pathogenesis and stability of asthma. NGF exacerbates airway inflammation in asthma by promoting smooth muscle cell proliferation and inducing the Th2 immune response. Activation of the Nrf2/HO-1 signaling pathway can exert a protective effect on the inflammatory response in bronchial asthma. However, the specific mechanism of this pathway in PM-involved asthmatic airway smooth muscle cells remains unclear. METHODS: Mice were sensitized and challenged with Ovalbumin (OVA) to establish an asthma model. They were then exposed to either PM, vitamin D or a combination of both, and inflammatory responses were observed. Including, acetylcholine stimulation at different concentrations measured airway hyperresponsiveness in mice. Bronchoalveolar lavage fluid (BALF) and serum were collected for TNF-α, IL-1ß, IL-6, and Nerve growth factor (NGF) analysis. Additionally, lung tissues underwent histopathological examination to observe alveolar structure and inflammatory cell infiltration. Specific ELISA kits were utilized to determine the levels of the inflammatory factors TNF-α, IL-1ß, IL-6, and Nerve growth factor (NGF). Nrf2/HO-1 signaling pathways were examined by western blot analysis. Meanwhile, we constructed a cell system with low HO-1 expression by lentiviral transfection of airway smooth muscle cells. The changes of Nrf2, HO-1, and NGF were observed after the treatment of OVA, PM, and Vit D were given. RESULTS: The in vivo results showed that vitamin D significantly alleviated pathological changes in lung tissue of PM-exposed mice models. Mechanismly, vitamin D decreased substantial inflammatory cell infiltration in lung tissue, as well as the number of inflammatory cells in BALF. Furthermore, vitamin D reduced the heightened inflammatory factors including of TNF-α, IL-1ß, IL-6, and NGF caused by PM exposure, and triggered the activity of nucleus Nrf2 and HO-1 in PM-exposed asthmatic mice. Notably, knockdown HO-1 weakens the Vitamin D- mediated inhibition to pollution toxicity in asthma. Importantly, in vitro experiments on OVA-stimulated mice airway smooth muscle cells, the results showed that OVA and PM, respectively, reduced Nrf2/HO-1 and increased NGF's expression, while vitamin D reversed the process. And in the HO-1 knockdown cell line of Lenti-si-HO-1 ASMCs, OVA and PM reduced Nrf2's expression, while HO-1 and NGF's expression were unchanged. CONCLUSIONS: The above results demastrate that vitamin D downregulated the inflammatory response and the expression of NGF by regulating the Nrf2/HO-1 signaling pathways in airway smooth muscle cells, thereby showing potent anti-inflammatory activity in asthma.

Radiofrequency ablation induces tumor cell dissemination in a mouse model of hepatocellular carcinoma.

BACKGROUND: We tested the hypothesis that radiofrequency ablation (RFA) for hepatocellular carcinoma (HCC) promotes tumor cell release and explored a method for reducing these effects. METHODS: A green fluorescent protein-transfected orthotopic HCC model was established in 99 nude mice. In vivo flow cytometry was used to monitor circulating tumor cell (CTC) dynamics. Pulmonary fluorescence imaging and pathology were performed to investigate lung metastases. First, the kinetics of CTCs during the periablation period and the survival rate of CTCs released during RFA were investigated. Next, mice were allocated to controls, sham ablation, or RFA with/without hepatic vessel blocking (ligation of the portal triads) for evaluating the postablation CTC level, lung metastases, and survival over time. Moreover, the kinetics of CTCs, lung metastases, and mice survival were evaluated for RFA with/without ethanol injection. Pathological changes in tumors and surrounding parenchyma after ethanol injection were noted. Statistical analysis included t-test, ANOVA, and Kaplan-Meier survival curves. RESULTS: CTC counts were 12.3-fold increased during RFA, and 73.7% of RFA-induced CTCs were viable. Pre-RFA hepatic vessel blocking prevented the increase of peripheral CTCs, reduced the number of lung metastases, and prolonged survival (all p ≤ 0.05). Similarly, pre-RFA ethanol injection remarkably decreased CTC release during RFA and further decreased lung metastases with extended survival (all p ≤ 0.05). Histopathology revealed thrombus formation in blood vessels after ethanol injection, which may clog tumor cell dissemination during RFA. CONCLUSION: RFA induces viable tumor cell dissemination, and pre-RFA ethanol injection may provide a prophylactic strategy to reduce this underestimated effect. RELEVANCE STATEMENT: RFA for HCC promotes viable tumor cell release during ablation, while ethanol injection can prevent RFA induced tumor cell release. KEY POINTS: • RFA induced the release of viable tumor cells during the ablation procedure in an animal model. • Hepatic vessel blocking can suppress tumor cells dissemination during RFA. • Ethanol injection can prevent RFA-induced tumor cell release, presumably because of the formation of thrombosis.

Turnover intention and coping strategies among older nursing assistants in China: a qualitative study.

Introduction: With the increasing aging population, older nursing assistants have made significant contributions to institutional eldercare. However, there is a high turnover rate among these workers, and it is crucial to address this issue and find ways to stabilize the workforce. This study aimed to explore the factors influencing turnover intention and coping strategies among older nursing assistants, in order to provide targeted assistance and guidance to reduce their intention to resign and ultimately lower the turnover rate. Methods: Qualitative research methods were employed to conduct semi-structured interviews with older nursing assistants in Changsha. The data obtained from these interviews were then analyzed using a phenomenological analysis approach and NVIVO (QSR International, Doncaster, Australia) software version 11.0. Results: It is found that several factors influence turnover intention among older nursing assistants. Which include work pay, work environment, professional identity, external motivation, and work pressure. Additionally, the coping strategies employed by these individuals in relation to their intention to resign include self-regulation, seeking support, self-improvement, and exploring motivation. Discussion: It is also evident from our study that reducing the turnover intention of older nursing assistants requires a collaborative effort from older adult care institutions, functional departments, and eldercare nursing assistants themselves. By addressing the factors influencing turnover intention and providing support and resources for coping strategies, we can work towards stabilizing the workforce and improving institutional eldercare.

A novel prognostic model based on AFP, tumor burden score and Albumin-Bilirubin grade for patients with hepatocellular carcinoma undergoing radiofrequency ablation.

PURPOSE: The aim of this study was to develop prognostic scores, including the tumor burden score (TBS) and albumin-bilirubin (ALBI) grade, for evaluating the outcomes of hepatocellular carcinoma (HCC) patients after radiofrequency ablation (RFA). MATERIALS AND METHODS: This retrospective study enrolled treatment-naïve HCC patients with BCLC 0-A who underwent RFA between January 2009 and December 2019. Regular follow-up was conducted after RFA to determine progression-free survival (PFS) and overall survival (OS). The patients were randomly allocated to the training or validation datasets in a 1:1 ratio. Preoperative prognostic scores were developed based on the results of multivariate analysis. The discriminatory ability of the scores was assessed using time-dependent AUC and compared with other models. RESULTS: Serum alpha-fetoprotein (AFP) level and TBS were identified as independent prognostic factors for PFS, while serum AFP, TBS, and ALBI were identified as independent prognostic factors for OS in HCC patients after RFA. The time-dependent AUCs of the AFP-TBS score for the 1-, 3-, and 5-year PFS were 0.651, 0.667, and 0.620, respectively, in the training set, and 0.657, 0.687, and 0.704, respectively, in the validation set. For the 1-, 3-, and 5-year OS, the time-dependent AUCs were 0.680, 0.712, and 0.666, respectively, in the training set, and 0.712, 0.706 and 0.726 in the validation set for the AFP-TBS-ALBI score (ATA). The C-indices and AIC demonstrated that the scores provided better clinical benefits compared to other models. CONCLUSION: The ATA/AT score, derived from clinical and objective laboratory variables, can assist in individually predicting the prognosis of HCC patients undergoing curative RFA.

Interpretable and accurate curve-fitting method for arterial pulse wave modeling and decomposition.

Arterial pulse waveforms contain a wealth of information about the cardiovascular system. There is a lack of physical meaning in the mathematical model of arterial pulse waves, while the physical model fails to offer individuality as too many assumptions are involved. In this article, we focus on promoting the interpretability of the arterial pulse mathematical model. The proposed method is based on newly developed 3-term fitting functions individualized by the physiological parameter assignment, which are the peak times of the reflected and dicrotic waves in a pulse. In this manner, the model allows decomposition of the pulse into sub-signals with clear physiological significance. With nearly 10,000 pulse fitting experiments, it is demonstrated that the proposed method outperforms the standard methods in fitting accuracy while providing parameters linked to hemodynamic characteristics and common clinical indices such as the peripheral augmentation index (pAI). The proposed method innovatively maintains the individuality and accuracy of mathematical models while improving the interpretability of their parameters. The applications of this newly developed method, which explicitly incorporates hemodynamic characteristics, are expected to be particularly valuable in future pulse wave decomposition studies.

Treatment outcomes after radiofrequency ablation in patients with non-B non-C hepatocellular carcinoma within Milan criteria: comparison with HBV-related hepatocellular carcinoma.

PURPOSE: This study aims to evaluate the treatment outcomes of radiofrequency ablation (RFA) for patients with non-B non-C hepatocellular carcinoma (HCC) (NBNC-HCC) within Milan criteria, as well as to compare them with those of patients with hepatitis B virus (HBV)-related HCC (HBV-HCC). METHODS: From January 2007 to February 2020, 303 patients with primary HCC who underwent RFA were retrospectively reviewed, including 259 patients with HBV-HCC (HBV-HCC group) and 44 patients with NBNC-HCC (NBNC-HCC group). The clinical characteristics and treatment survivals were evaluated and compared. Moreover, the propensity score matching was used to reduce selection bias. RESULTS: A significantly lower proportion of cirrhosis was observed in the NBNC-HCC group (p = .048). Before propensity score matching, local tumor progression, disease-free survival, and overall survival after RFA showed no significant differences between the two groups (all p > .05). After matching, the overall survival rates in the NBNC-HCC group were significantly better than those in the HBV-HCC group (p = .042). Moreover, for patients with NBNC-HCC, tumor size (hazard ratio = 8.749, 95% confidence interval, 1.599-47.849; p = .012) was the only independent predictor of local tumor progression. CONCLUSIONS: Patients with NBNC-HCC within the Milan criteria after RFA had better long-term survival than patients with HBV-HCC, although larger, prospective and multicenter trials are required to validate these results.

Toward Intrinsic Adversarial Robustness Through Probabilistic Training.

Modern deep neural networks have made numerous breakthroughs in real-world applications, yet they remain vulnerable to some imperceptible adversarial perturbations. These tailored perturbations can severely disrupt the inference of current deep learning-based methods and may induce potential security hazards to artificial intelligence applications. So far, adversarial training methods have achieved excellent robustness against various adversarial attacks by involving adversarial examples during the training stage. However, existing methods primarily rely on optimizing injective adversarial examples correspondingly generated from natural examples, ignoring potential adversaries in the adversarial domain. This optimization bias can induce the overfitting of the suboptimal decision boundary, which heavily jeopardizes adversarial robustness. To address this issue, we propose Adversarial Probabilistic Training (APT) to bridge the distribution gap between the natural and adversarial examples via modeling the latent adversarial distribution. Instead of tedious and costly adversary sampling to form the probabilistic domain, we estimate the adversarial distribution parameters in the feature level for efficiency. Moreover, we decouple the distribution alignment based on the adversarial probability model and the original adversarial example. We then devise a novel reweighting mechanism for the distribution alignment by considering the adversarial strength and the domain uncertainty. Extensive experiments demonstrate the superiority of our adversarial probabilistic training method against various types of adversarial attacks in different datasets and scenarios.

Cross-Camera Trajectories Help Person Retrieval in a Camera Network.

We are concerned with retrieving a query person from multiple videos captured by a non-overlapping camera network. Existing methods often rely on purely visual matching or consider temporal constraints but ignore the spatial information of the camera network. To address this issue, we propose a pedestrian retrieval framework based on cross-camera trajectory generation that integrates both temporal and spatial information. To obtain pedestrian trajectories, we propose a novel cross-camera spatio-temporal model that integrates pedestrians' walking habits and the path layout between cameras to form a joint probability distribution. Such a cross-camera spatio-temporal model can be specified using sparsely sampled pedestrian data. Based on the spatio-temporal model, cross-camera trajectories can be extracted by the conditional random field model and further optimised by restricted non-negative matrix factorization. Finally, a trajectory re-ranking technique is proposed to improve the pedestrian retrieval results. To verify the effectiveness of our method, we construct the first cross-camera pedestrian trajectory dataset, the Person Trajectory Dataset, in real surveillance scenarios. Extensive experiments verify the effectiveness and robustness of the proposed method.

Wavelet-Guided Promotion-Suppression Transformer for Surface-Defect Detection.

Surface-defect detection aims to accurately locate and classify defect areas in images via pixel-level annotations. Different from the objects in traditional image segmentation, defect areas comprise a small group of pixels with random shapes, characterized by uncommon textures and edges that are inconsistent with the normal surface patterns of industrial products. This task-specific knowledge is hardly considered in the current methods. Therefore, we propose a two-stage "promotion-suppression" transformer (PST) framework, which explicitly adopts the wavelet features to guide the network to focus on the detailed features in the images. Specifically, in the promotion stage, we propose the Haar augmentation module to improve the backbone's sensitivity to high-frequency details. However, the background noise is inevitably amplified as well because it also constitutes high-frequency information. Therefore, a quadratic feature-fusion module (QFFM) is proposed in the suppression stage, which exploits the two properties of noise: independence and attenuation. The QFFM analyzes the similarities and differences between noise and defect features to achieve noise suppression. Compared with the traditional linear-fusion approach, the QFFM is more sensitive to high-frequency details; thus, it can afford highly discriminative features. Extensive experiments are conducted on three datasets, namely DAGM, MT, and CRACK500, which demonstrate the superiority of the proposed PST framework.

Ultrasound-Guided Human Islet Transplantation: Safety, Feasibility, and Efficacy Analysis.

RATIONALE AND OBJECTIVES: We aimed to analyze the safety, feasibility, and efficacy of human islet transplantation (IT) using ultrasound (US) throughout the entire procedure. MATERIALS AND METHODS: A total of 22 recipients (18 males; mean age 42.6 ± 17.5years) with 35 procedures were retrospective included. Under US guidance, percutaneous transhepatic portal catheterization was successfully performed through a right-sided transhepatic approach, and islets were infused into the main portal vein. Color Doppler and contrast-enhanced ultrasound were used to guide the procedure and monitor the complications. After infusion of the islet mass, the access track was embolized by embolic material. If hemorrhage persisted, US-guided radiofrequency ablation (RFA) was performed to stop bleeding. Factors that could affect the complication were analyzed. After transplantation, primary graft function was evaluated with a ß-score 1month after the last islet infusion. RESULTS: The technical success rates were 100% with a single puncture attempt. Six (17.1%) abdominal bleeding episodes were immediately stopped by US-guided RFA. No portal vein thrombosis were encountered. Dialysis (OR (Odd Ratio): 32.0; 95% CI: 1.561-656.054; and P = .025) was identified as a significant factor associated with bleeding. Primary graft function was optimal in eight patients (36.4%), suboptimal in 13 patients (59.1%), and poor in one patient (4.5%). CONCLUSION: In conclusion, whole-procedure US-guided IT is a safe, feasible, and effective method for diabetes. Complications are either self-limiting or manageable with noninvasive treatment.

Predicting symptomatic post-hepatectomy liver failure in patients with hepatocellular carcinoma: development and validation of a preoperative nomogram.

OBJECTIVE: To develop and validate a nomogram based on liver stiffness (LS) for predicting symptomatic post-hepatectomy (PHLF) in patients with hepatocellular carcinoma (HCC). METHODS: A total of 266 patients with HCC were enrolled prospectively from three tertiary referral hospitals from August 2018 to April 2021. All patients underwent preoperative laboratory examination to obtain parameters of liver function. Two-dimensional shear wave elastography (2D-SWE) was performed to measure LS. Three-dimensional virtual resection obtained the different volumes including future liver remnant (FLR). A nomogram was developed by using logistic regression and determined by receiver operating characteristic (ROC) curve analysis and calibration curve analysis, which was validated internally and externally. RESULTS: A nomogram was constructed with the following variables: FLR ratio (FLR of total liver volume), LS greater than 9.5 kPa, Child-Pugh grade, and the presence of clinically significant portal hypertension (CSPH). This nomogram enabled differentiation of symptomatic PHLF in the derivation cohort (area under curve [AUC], 0.915), internal fivefold cross-validation (mean AUC, 0.918), internal validation cohort (AUC, 0.876) and external validation cohort (AUC, 0.845). The nomogram also showed good calibration in the derivation, internal validation, and external validation cohorts (Hosmer-Lemeshow goodness-of-fit test, p = 0.641, p = 0.06, and p = 0.127, respectively). Accordingly, the safe limit of the FLR ratio was stratified using the nomogram. CONCLUSION: An elevated level of LS was associated with the occurrence of symptomatic PHLF in HCC. A preoperative nomogram integrating LS, clinical and volumetric features was useful in predicting postoperative outcomes in patients with HCC, which might help surgeons in the management of HCC resection. CLINICAL RELEVANCE STATEMENT: A serial of the safe limit of the future liver remnant was proposed by a preoperative nomogram for hepatocellular carcinoma, which might help surgeons in 'how much remnant is enough in liver resection'. KEY POINTS: • An elevated liver stiffness with the best cutoff value of 9.5 kPa was associated with the occurrence of symptomatic post-hepatectomy liver failure in hepatocellular carcinoma. • A nomogram based on both quality (Child-Pugh grade, liver stiffness, and portal hypertension) and quantity of future liver remnant was developed to predict symptomatic post-hepatectomy liver failure for HCC, which enabled good discrimination and calibration in both derivation and validation cohorts. • The safe limit of future liver remnant volume was stratified using the proposed nomogram, which might help surgeons in the management of HCC resection.

Pose Guided Person Image Generation via Dual-task Correlation and Affinity Learning.

Pose Guided Person Image Generation (PGPIG) is the task of transforming a person's image from the source pose to a target pose. Existing PGPIG methods often tend to learn an end-to-end transformation between the source image and the target image, but do not seriously consider two issues: 1) the PGPIG is an ill-posed problem, and 2) the texture mapping requires effective supervision. In order to alleviate these two challenges, we propose a novel method by incorporating Dual-task Pose Transformer Network and Texture Affinity learning mechanism (DPTN-TA). To assist the ill-posed source-to-target task learning, DPTN-TA introduces an auxiliary task, i.e., source-to-source task, by a Siamese structure and further explores the dual-task correlation. Specifically, the correlation is built by the proposed Pose Transformer Module (PTM), which can adaptively capture the fine-grained mapping between sources and targets and can promote the source texture transmission to enhance the details of the generated images. Moreover, we propose a novel texture affinity loss to better supervise the learning of texture mapping. In this way, the network is able to learn complex spatial transformations effectively. Extensive experiments show that our DPTN-TA can produce perceptually realistic person images under significant pose changes. Furthermore, our DPTN-TA is not limited to processing human bodies but can be flexibly extended to view synthesis of other objects, i.e., faces and chairs, outperforming the state-of-the-arts in terms of both LPIPS and FID. Our code is available at: https://github.com/PangzeCheung/Dual-task-Pose-Transformer-Network.