One-stage posterior transpedicular debridement, hemi-interbody and unilateral-posterior bone grafting, and instrumentation for the treatment of thoracic spinal tuberculosis: a retrospective study.

PURPOSE: To investigate the clinical efficacy and feasibility of the surgical treatment of thoracic spinal tuberculosis using one-stage posterior instrumentation, transpedicular debridement, and hemi-interbody and unilateral posterior bone grafting. METHODS: Fifty-six patients with thoracic spinal tuberculosis who underwent surgery performed by a single surgeon between September 2009 and August 2020 were enrolled in this study. Based on data from the erythrocyte sedimentation rate (ESR), Visual Analog Scale (VAS), and Cobb angle before surgery, after surgery, and at the most recent follow-up, clinical effectiveness was assessed using statistical analysis. The variables investigated included operating time, blood loss, complications, neurological function, and hemi-interbody fusion. RESULTS: None of the patients experienced significant surgery-associated complications. At the last follow-up, 23 of the 25 patients (92%) with neurological impairment showed improvement. The thoracic kyphotic angle was significantly decreased from 24.1 ± 9.9° to 13.4 ± 8.6° after operation (P < 0.05), and the angle was 14.44 ± 8.8° at final follow-up (P < 0.05). The Visual Analog Scale significantly decreased from 6.7 ± 1.4 preoperatively to 2.3 ± 0.8 postoperatively (P < 0.05) and finally to 1.2 ± 0.7 at the last follow-up (P < 0.05). Bone fusion was confirmed in 56 patients at 3-6 months postoperatively. CONCLUSIONS: One-stage posterior transpedicular debridement, hemi-interbody and unilateral posterior bone grafting, and instrumentation are effective and feasible treatment methods for thoracic spinal tuberculosis.

Revealing the Dominance of the Dissolution-Deposition Mechanism in Aqueous Zn-MnO2 Batteries.

Zn-MnO2 batteries have attracted extensive attention for grid-scale energy storage applications, however, the energy storage chemistry of MnO2 in mild acidic aqueous electrolytes remains elusive and controversial. Using α-MnO2 as a case study, we developed a methodology by coupling conventional coin batteries with customized beaker batteries to pinpoint the operating mechanism of Zn-MnO2 batteries. This approach visually simulates the operating state of batteries in different scenarios and allows for a comprehensive study of the operating mechanism of aqueous Zn-MnO2 batteries under mild acidic conditions. It is validated that the electrochemical performance can be modulated by controlling the addition of Mn2+ to the electrolyte. The method is utilized to systematically eliminate the possibility of Zn2+ and/or H+ intercalation/de-intercalation reactions, thereby confirming the dominance of the MnO2 /Mn2+ dissolution-deposition mechanism. By combining a series of phase and spectroscopic characterizations, the compositional, morphological and structural evolution of electrodes and electrolytes during battery cycling is probed, elucidating the intrinsic battery chemistry of MnO2 in mild acid electrolytes. Such a methodology developed can be extended to other energy storage systems, providing a universal approach to accurately identify the reaction mechanism of aqueous aluminum-ion batteries as well.

Analysis of the 2p-manifold population distribution in a diode-pumped metastable Ar laser.

The complex excited energy levels in the diode-pumped metastable Ar laser may induce harmful effects in laser cycling. Significantly, the influence of the population distribution in 2p energy levels on the laser performance is unclear yet. In this work, the absolute populations in all the 2p states were measured online by the simultaneous applications of tunable diode laser absorption spectroscopy and optical emission spectroscopy. The results showed that most atoms were populated to the 2p8, 2p9, and 2p10 levels while lasing, and the majority of the 2p9 population was efficiently transferred to the 2p10 level with the aid of helium, which was beneficial for the laser performance.

Regulation of the Oxidase Mimetic Activity of Ceria Nanoparticles by Buffer Composition.

Ceria nanoparticles (CNPs) are important typical nanozymes with multiple enzyme mimetic activities, which could facilitate the oxidation of organic dyes in acidic conditions, because of the oxidase mimetic activity. Usually, the regulation of oxidase mimetic activity is focused on the adjustment of the structure, morphology, composition, surface, and other factors of nanozymes. However, the influence of the surrounding environment is not considered, which is very important during the reaction process. In this work, the oxidase mimetic activity of CNPs in buffer solutions including citric acid, acetic acid and glycine buffer solutions was investigated, with the results that carboxyl group in buffer solution could adsorb the CNPs on the surface to promote the oxidase mimetic activity. Due to the chelation with the cerium ion, the enhancement is more significant by molecules with polycarboxylic groups, and the enhancement is more efficient by carboxyl molecules in buffer solution, compared with the modification of the carboxyl groups on the surface, because of easier operation and smaller steric hindrance. From the viewpoint of increasing the oxidase mimetic activity of CNPs, the work is expected to provide references for the selection of the reaction systems to optimize the oxidase mimetic activity in bio-detection applications.

Acute effects of foam rolling and dynamic stretching on angle-specific change of direction ability, flexibility and reactive strength in male basketball players.

The purpose of our study was to determine the acute effects of dynamic stretching (DS), foam rolling (FR) and foam rolling combined with dynamic stretching (Combo) protocols on angle-specific change of direction (COD) ability, drop jump (DJ) performance and flexibility. Using a counterbalance crossover study design, eleven male basketball collegiate players (20.7 ± 0.6 years) were randomly assigned to one of the four protocols - control (CON), DS, FR, Combo - for each session, for a total of four sessions. A more aggressive foam cylinder with raised nodules, which is thought to be effective in stimulating the deep layer of muscle tissue, was used to observe for changes in their performance during sit and reach (SAR), DJ and COD tasks in 45 and 180 degrees. One-way repeated measures ANOVA was used to identify differences of each variable separately between interventions. The SAR after three interventions compared to the CON was significantly improved (F (3,30) = 5.903, P = 0.003, η2 = 0.371). In the 505 test, both limbs failed to show a significant improvement in COD deficit. The non-dominant limb showed a significant improvement of 6.4% after FR when performing the Y-shaped agility (F (3,30) = 4.962, P = 0.0065 < 0.05, η2 = 0.332). In the DJ, the reactive strength index and contact time changed significantly by 17.5% and -17.5% (η2 = 0.518, η2 = 0.571), respectively, immediately after FR. The current research suggested that FR may have an enhancing effect on COD speed in a 45° cutting task and neuromuscular function, while having the potential to improve non-dominant limb deficits in both COD tasks. In contrast, the Combo warm-up protocol did not produce a cumulative effect, suggesting the need for coaches to remain cautious about excessive warm-up duration.

Demonstration of a diode-pumped plasma jet-type rare gas laser.

The diode-pumped metastable rare gas laser (DPRGL) is showing potential for high-power operation. A key issue in developing this concept is to produce high-density metastables in a large volume. To achieve this goal, we propose a new, to the best of our knowledge, architecture by extracting laser power from a diode-pumped plasma jet. In this scheme, the discharge and gain regions are separated, avoiding the negative effects of discharges in confined regions. A diode-pumped plasma jet-type Ar laser is demonstrated with 466-mW output and 33% slope efficiency. The gain volume can be increased with multi-jets, providing a better scaling potential for the DPRGL system.

Revealing kinetics of a diode-pumped metastable Ar laser in pulsed and CW lasing.

We have experimentally investigated the kinetics of a diode-pumped metastable Ar laser by simultaneously monitoring the population evolution of 1s5 and 1s4 states during lasing. A comparison between the two cases with the pump laser on and off revealed the cause for the transition from pulsed to CW lasing. The depletion of 1s5 atoms was responsible for the pulsed lasing phenomenon, while increasing the duration and density of 1s5 atoms resulted in CW lasing. Furthermore, population accumulation of the 1s4 state was observed.

Bacterial succession and the dynamics of flavor compounds in the Huangjiu fermented from corn.

Huangjiu is one of the Chinese unique and traditional liquor. Corn is a kind of high yield crop in China, which has the characteristics of wide distribution, low price and high starch content. Fermenting Huangjiu with corn not only enrich Huangjiu types, but also opens up a new way for the utilization of corn. The flavor compounds and microorganisms in corn wine fermentation were studied in this study. A total of 98 volatile compounds and 8 kinds of organic acids were detected. Bacillus, Weissella, Streptomyces, Aeromonas and Blautia were the dominant bacteria. The correlation analysis between flavor compounds and bacteria showed that there were 557 correlations between major flavor compounds and bacteria. Among them, Lactococcus, Virgibacillus, Sphingobacterium and Sporolactobacillus were dominant genus of flavor producing bacteria. This study may reveal the changing rule of bacteria in Huangjiu, predict the relationship between metabolites and bacteria. In addition, this study expanded the application of corn and increased the variety of Huangjiu.

Neurotransmitter system aberrations in patients with drug addiction.

Drug dependence may affect the neurotransmitter system levels in the human body. This study recruited 113 healthy control subjects, 118 heroin-dependent patients and 118 methamphetamine-dependent patients and examined the serum 5-HT, dopamine, glutamate and norepinephrine levels in the 349 volunteers. ELISA assays demonstrated that the serum 5-HT levels were significantly reduced in the drug-dependent patients, whereas the serum dopamine and glutamate levels were both significantly increased in the drug-dependent patients when compared with control subjects. In contrast, the norepinephrine levels did not exhibit a significant difference between the drug-dependent and control subjects. We also used qRT-PCR to analyze the transcriptional expression levels of 5-HT1A, 5-HT1B, dopmaine-D1 and dopamine-D2 receptors in the blood of drug-dependent patients and controls, and the results show that only 5-HT1B receptor levels were dysfunctional in the heroin abusers. In addition, our results suggest that serum 5-HT, dopamine, and glutamate levels had the potential to differ between drug abusers and controls, and combining those three potential biomarkers provided an accurate means to differentiate between the drug-dependent and control subjects. Taken together, our study reveals a differential profile of neurotransmitters in the heroin-dependent patients and methamphetamine-dependent patients, and this revelation may contribute to understanding the pathophysiology of drug addiction.

Comparative transcriptomic analysis of hypothalamus-pituitary-liver axis in bighead carp (Hypophthalmichthys nobilis) with differential growth rate.

BACKGROUND: Growth rate is one of the most important features for aquaculture species and deciphering its regulation mechanism has great significance both in genetics and in economics. Hypothalamus-pituitary growth axis (HP growth axis) or neuro-endocrine axis plays a vital role in growth regulation in different aquaculture animals. RESULTS: In this study, the HP and liver transcriptomes of two female groups (H and L) with phenotypically extreme growth rate were sequenced using RNA-Seq. A total of 30,524 and 22,341 genes were found expressed in the two tissues, respectively. The average expression levels for the two tissues were almost the same, but the median differed significantly. A differential expression analysis between H and L groups identified 173 and 204 differentially expressed genes (DEGs) in HP and liver tissue, respectively. Pathway analysis revealed that DEGs in HP tissue were enriched in regulation of cell proliferation and angiogenesis while in liver tissue these genes were overrepresented in sterol biosynthesis and transportation. Genomic overlapping analyses found that 4 and 5 DEGs were within growth-related QTL in HP and liver tissue respectively. A deeper analysis of these 9 genes revealed 3 genes were functionally linked to the trait of interest. The expression of 2075 lncRNAs in HP tissue and 1490 in liver tissue were also detected, and some of lncRNAs were highly expressed in the two tissues. CONCLUSIONS: Above all, the results of the present study greatly contributed to the knowledge of the regulation of growth and then assisted the design of new selection strategies for bighead carp with improved growth-related traits.

Astragaloside IV protects against diabetic nephropathy via activating eNOS in streptozotocin diabetes-induced rats.

BACKGROUND: Astragaloside IV (AS-IV) was reported to play a role in improving diabetic nephropathy (DN), however, the underlying mechanisms still remain unclear. The aim of the present study is to investigate whether AS-IV ameliorates DN via the regulation of endothelial nitric oxide synthase (eNOS). METHODS: DN model was induced in Sprague-Dawley (SD) male rats by intraperitoneal injection of 65 mg/kg streptozotocin (STZ). Rats in the AS-IV treatment group were orally gavaged with 5 mg/kg/day or 10 mg/kg/day AS-IV for eight consecutive weeks. Body weight, blood glucose, blood urea nitrogen (BUN), Serum creatinine (Scr), proteinuria and Glycosylated hemoglobin (HbA1c) levels were measured. Hematoxylin-Eosin (HE) and Periodic Acid-Schiff (PAS) staining were used to detect the renal pathology. The apoptosis status of glomerular cells was measured by TUNEL assay. The phosphorylation and acetylation of eNOS were detected by western blot. The effects of AS-IV on high-glucose (HG)-induced apoptosis and eNOS activity were also investigated in human renal glomerular endothelial cells (HRGECs) in vitro. RESULTS: Treatment with AS-IV apparently reduced DN symptoms in diabetic rats, as evidenced by reduced BUN, Scr, proteinuria, HbA1c levels and expanding mesangial matrix. AS-IV treatment also promoted the synthesis of nitric oxide (NO) in serum and renal tissues and ameliorated the phosphorylation of eNOS at Ser 1177 with decreased eNOS acetylation. Moreover, HG-induced dysfunction of HRGECs including increased cell permeability and apoptosis, impaired eNOS phosphorylation at Ser 1177, and decreased NO production, were all reversed by AS-IV treatment. CONCLUSIONS: These novel findings suggest that AS-IV ameliorates functional abnormalities of DN through inhibiting acetylation of eNOS and activating its phosphorylation at Ser 1177. AS-IV could be served as a potential therapeutic drug for DN.

Innovative method for the enrichment of high-polarity bioactive molecules present at low concentrations in complex matrices.

Ginsenoside Rg1 is a valuable bioactive molecule but its high polarity and low concentration in complex mixtures makes it a challenge to separate Ginsenoside Rg1 from other saponins with similar structures, resulting in low extraction efficiency. The successful development of effective Rg1 molecularly imprinted polymers that exhibit high selectivity and adsorption may offer an improved method for the enrichment of active compounds. In this work, molecularly imprinted polymers were prepared with two different methods, precipitation polymerization or surface imprinted polymerization. Comparison of the adsorption abilities showed higher adsorption of the surface molecularly imprinted polymers prepared by surface imprinted polymerization, 46.80 mg/g, compared to the 27.74 mg/g observed for the molecularly imprinted polymers prepared by precipitation polymerization. Therefore, for higher adsorption of the highly polar Rg1, surface imprinted polymerization is a superior technique to make Rg1 molecularly imprinted polymers. The prepared surface molecularly imprinted polymers were tested as a solid-phase extraction column to directionally enrich Rg1 and its analogues from ginseng tea and total ginseng extracts. The column with surface molecularly imprinted polymers showed higher enrichment efficiency and better selectivity than a C18 solid-phase extraction column. Overall, a new, innovative method was developed to efficiently enrich high-polarity bioactive molecules present at low concentrations in complex matrices.

Genome-wide identification and expression analysis of SBP-box gene family in Sorghum bicolor L.

SQUAMOSA PROMOTER BINDING PROTEIN-box (SBP-box) family genes encoding plant-specific transcription factors are involved in many aspects of crop genetic improvement such as yield, plant-type and stress-resistance. The SBP-box gene family have important practical applications. In this study, 18 SBP-box genes were identified from the reference genome of sorghum (Sorghum bicolor L.) using bioinformatics. These genes distributed on nine chromosomes while eight of them located in the segmental duplication region. Phylogenetic reconstruction resulted in six subfamilies of SBP-box genes, among which SbSBP12, SbSBP3 and SbSBP15 are orthologous to ZmLG1, ZmTGA1 and ZmUB2/3 in corn, respectively. RNA-seq data analysis indicated that SbSBP-box genes show the highest expression level in primordial inflorescences. Moreover, SbSBP9 and SbSBP17 exhibited a tissue specific expression in primordial inflorescences. The expression levels of SbSBP5, SbSBP8 and SbSBP18 were increased in response to exogenous ABA and PEG,indicating that SbSBP-box genes are involved in the defense response against abiotic stresses in sorghum. This research provides references for cloning important genes in SbSBP-box gene family. Genes identified in this study could be considered as candidate genes for genetic improvement of sorghum.

[Neuro-protective effects and mechanism of Terminalia chebula extract HZ4 on cerebral infarction models].

To study the effect and possible molecular mechanisms of Terminalia chebula extract HZ4 on focal cerebral infarction in rats, 90 healthy male SD rats were randomly divided into sham-operation group, model group, T. chebula extract HZ4 high dose, middle dose and low dose groups (80, 40, 20 mg•kg ⁻¹â€¢d ⁻¹, ig) and positive control group (Panax notoginseng saponins, PNS 30 mg•kg ⁻¹â€¢d ⁻¹, ig). The focal cerebral infarction models were established by photochemical method. After the rats were administered for 7 consecutive days, neurogenic behavior rating of these rats was done by balance beam test and foot fault test. The cells morphological changes of penumbra in focal cerebral infarction were investigated by HE staining method; the infarct volume was detected by TTC staining. The expression levels of ß-catenin and cyclin D1, the key node genes in Wnt signaling pathway of the focal penumbra tissues were detected via RT-PCR. The results showed that, as compared with the model group, behavioural indicators were improved significantly in the rats of administration groups, and the infarct volume and pathological changes of penumbra tissues were also improved at the same time. Compared with the model group, the expression levels of ß-catenin and cyclin D1 in Wnt signaling pathway were significantly up-regulated in administration groups(P<0.01). This study first confirmed that T. chebula extract HZ4 can decrease infarct volume, improve the sport ability score, and promote rehabilitation of model animals. In addition, it could significant up-regulated the expression levels of ß-catenin and cyclin D1, and the mechanism may be associated with Wnt signaling pathway. The study is innovative to a certain extent.

[Analysis on signaling pathway network of proliferation of neural stem cells].

Neural stem cells in brains have capacities of proliferation and differentiation, which is very critical to rebuild the cerebral cortex functions. Therefore, it is of great importance to find key targets and network pathways that regulate the proliferation of neural stem cells, which is also a pressing problem in the medical circle. With the Notch pathway as the core of the network, this paper summarized the advance of the bimolecular network system composed of Wnt, Shh, EGFR, cytokines and Notch signal, and analyzed such key nodes as Notch receptor, CBF1, NICD, Hesl, which may become potential targets of new-type drugs in the future. With the multi-component, multi-target, multi-lever characteristics, traditional Chinese medicines have many common grounds with the network pharmacology. The active component groups or active ingredients in traditional Chinese medicines are one of the material bases for showing their network pharmacological effect, which is worth exploring. This paper aims to provide a new strategy for the treatment of neurodegenerative disease and nerve injury with traditional Chinese medicines.

Distributed deep reinforcement learning based on bi-objective framework for multi-robot formation.

Improving generalization ability in multi-robot formation can reduce repetitive training and calculation. In this paper, we study the multi-robot formation problem with the ability to generalize the target position. Since the generalization ability of neural network is directly proportional to spatial dimension, we adopt the strategy of using different networks to solve different objectives, so that the network learning can focus on the learning of one objective to obtain better performance. In addition, this paper presents a distributed deep reinforcement learning method based on soft actor-critic algorithm for solving multi-robot formation problem. At the same time, the formation evaluation assignment function is designed to adapt to distributed training. Compared with the original algorithm, the improved algorithm can get higher reward cumulative values. The experimental results show that the proposed algorithm can better maintain the desired formation in the moving process, and the rotation design in the reward function makes the multi-robot system have better flexibility in formation. The comparison of control signal curve shows that the proposed algorithm is more stable. At the end of the experiments, the universality of the proposed algorithm in formation maintenance and formation variations is demonstrated.

Cytisine-N-methylene-(5,7,4'-trihydroxy)- isoflavone ameliorates ischemic stroke-induced brain injury in mouse by regulating the oxidative stress and BDNF-Trkb/Akt pathway.

BACKGROUND: A novel compound Cytisine-N-methylene-(5,7,4'-trihydroxy)- isoflavone (LY01) found in the Sophora alopecuroides L is a neuroprotective agent. However, the effect and potential mechanism of LY01 treatment for ischemic stroke (IS) have not been fully elucidated. AIM OF THE STUDY: The aim of this study is to demonstrate whether LY01 can rescue ischemic stroke-induced brain injury and oxygen-glucose deprivation/reperfusion (OGD/R). RESULTS: Our results show that intragastric administration of LY01 improves ischemic stroke behaviors in mice, as demonstrated by neurological score, infarct volume, cerebral water content, rotarod test for activity. Compared with the model group, the ginkgo biloba extract (EGb) and LY01 reversed the neurological score, infarct volume, cerebral water content, rotarod test in model mice. Further analysis showed that the LY01 rescued oxidative stress in the model mice, which was reflected in the increased levels of catalase, superoxide dismutase, total antioxidant capacity and decreased levels of malondialdehyde in the serum of the model mice. Moreover, the expression of the brain-derived neurotrophic factor brain-derived neurotrophic factor (BDNF), phosphorylated protein kinase B (p-Akt), Bax, Bcl-2, (p)-tropomysin related kinase B (p-Trkb) was restored and the expression of Bax, glial fibrillary acidic protein (GFAP) in the brains of the model mice was inhibited through LY01 treatment. In the polymerase chain reaction (PCR) data, after giving LY01, the expression in the brains of model mice was that, IL-10 increased and IL-1ß, Bax, Bcl-2 decreased. Furthermore, the results indicated that LY01 improved cell viability, reactive oxygen species content, and mitochondrial membrane potential dissipation induced by OGD/R in primary culture of rat cortical neurons. Bax and caspase-3 activity was upregulated compared to the before after treatment with LY01. CONCLUSIONS: Our study suggests that LY01 reversed ischemic stroke by reducing oxidative stress and activating the BDNF-TrkB/Akt pathway and exerted a neuroprotective action against OGD/R injury via attenuation, a novel approach was suggested to treat ischemic stroke. Our observations justify the traditional use of LY01 for a treatment of IS in nervous system.

Predefined-time distributed optimization and anti-disturbance control for nonlinear multi-agent system with neural network estimator: A hierarchical framework.

This paper addresses the predefined-time distributed optimization of nonlinear multi-agent system using a hierarchical control approach. Considering unknown nonlinear functions and external disturbances, we propose a two-layer hierarchical control framework. At the first layer, a predefined-time distributed estimator is employed to produce optimal consensus trajectories. At the second layer, a neural-network-based predefined-time disturbance observer is introduced to estimate the disturbance, with neural networks used to approximate the unknown nonlinear functions. A neural-network-based anti-disturbance sliding mode control mechanism is presented to ensure that the system trajectories can track the optimal trajectories within a predefined time. The feasibility of this hierarchical control framework is verified by utilizing the Lyapunov method. Numerical simulations are conducted separately using models of robotic arms and mobile robots to validate the effectiveness of the proposed method.

Geometric Understanding of Discriminability and Transferability for Visual Domain Adaptation.

To overcome the restriction of identical distribution assumption, invariant representation learning for unsupervised domain adaptation (UDA) has made significant advances in computer vision and pattern recognition communities. In UDA scenario, the training and test data belong to different domains while the task model is learned to be invariant. Recently, empirical connections between transferability and discriminability have received increasing attention, which is the key to understand the invariant representations. However, theoretical study of these abilities and in-depth analysis of the learned feature structures are unexplored yet. In this work, we systematically analyze the essentials of transferability and discriminability from the geometric perspective. Our theoretical results provide insights into understanding the co-regularization relation and prove the possibility of learning these abilities. From methodology aspect, the abilities are formulated as geometric properties between domain/cluster subspaces (i.e., orthogonality and equivalence) and characterized as the relation between the norms/ranks of multiple matrices. Two optimization-friendly learning principles are derived, which also ensure some intuitive explanations. Moreover, a feasible range for the co-regularization parameters is deduced to balance the learning of geometric structures. Based on the theoretical results, a geometry-oriented model is proposed for enhancing the transferability and discriminability via nuclear norm optimization. Extensive experiment results validate the effectiveness of the proposed model in empirical applications, and verify that the geometric abilities can be sufficiently learned in the derived feasible range.

Cross-site prognosis prediction for nasopharyngeal carcinoma from incomplete multi-modal data.

Accurate prognosis prediction for nasopharyngeal carcinoma based on magnetic resonance (MR) images assists in the guidance of treatment intensity, thus reducing the risk of recurrence and death. To reduce repeated labor and sufficiently explore domain knowledge, aggregating labeled/annotated data from external sites enables us to train an intelligent model for a clinical site with unlabeled data. However, this task suffers from the challenges of incomplete multi-modal examination data fusion and image data heterogeneity among sites. This paper proposes a cross-site survival analysis method for prognosis prediction of nasopharyngeal carcinoma from domain adaptation viewpoint. Utilizing a Cox model as the basic framework, our method equips it with a cross-attention based multi-modal fusion regularization. This regularization model effectively fuses the multi-modal information from multi-parametric MR images and clinical features onto a domain-adaptive space, despite the absence of some modalities. To enhance the feature discrimination, we also extend the contrastive learning technique to censored data cases. Compared with the conventional approaches which directly deploy a trained survival model in a new site, our method achieves superior prognosis prediction performance in cross-site validation experiments. These results highlight the key role of cross-site adaptability of our method and support its value in clinical practice.