Machine learning-driven source identification and ecological risk prediction of heavy metal pollution in cultivated soils.

To overcome challenges in assessing the impact of environmental factors on heavy metal accumulation in soil due to limited comprehensive data, our study in Yangxin County, Hubei Province, China, analyzed 577 soil samples in combination with extensive big data. We used machine learning techniques, the potential ecological risk index, and the bivariate local Moran's index (BLMI) to predict Cr, Pb, Cd, As, and Hg concentrations in cultivated soil to assess ecological risks and identify pollution sources. The random forest model was selected for its superior performance among various machine learning models, and results indicated that heavy metal accumulation was substantially influenced by environmental factors such as climate, elevation, industrial activities, soil properties, railways, and population. Our ecological risk assessment highlighted areas of concern, where Cd and Hg were identified as the primary threats. BLMI was used to analyze spatial clustering and autocorrelation patterns between ecological risk and environmental factors, pinpointing areas that require targeted interventions. Additionally, redundancy analysis revealed the dynamics of heavy metal transfer to crops. This detailed approach mapped the spatial distribution of heavy metals, highlighted the ecological risks, identified their sources, and provided essential data for effective land management and pollution mitigation.

Regulation and mechanistic insights into tensile strain in mesenchymal stem cell osteogenic differentiation.

Bone marrow mesenchymal stem cells (BMSCs) are integral to bone remodeling and homeostasis, as they are capable of differentiating into osteogenic and adipogenic lineages. This differentiation is substantially influenced by mechanosensitivity, particularly to tensile strain, which is a prevalent mechanical stimulus known to enhance osteogenic differentiation. This review specifically examines the effects of various cyclic tensile stress (CTS) conditions on BMSC osteogenesis. It delves into the effects of different loading devices, magnitudes, frequencies, elongation levels, dimensionalities, and coculture conditions, providing a comparative analysis that aids identification of the most conducive parameters for the osteogenic differentiation of BMSCs. Subsequently, this review delineates the signaling pathways activated by CTS, such as Wnt/ß-catenin, BMP, Notch, MAPK, PI3K/Akt, and Hedgehog, which are instrumental in mediating the osteogenic differentiation of BMSCs. Through a detailed examination of these pathways, this study elucidates the intricate mechanisms whereby tensile strain promotes osteogenic differentiation, offering valuable guidance for optimizing therapeutic strategies aimed at enhancing bone regeneration.

The tumor-stroma ratio in giant cell tumor of bone: associations with the immune microenvironment and responsiveness to denosumab treatment.

BACKGROUND: Currently, there is limited understanding regarding the clinical significance of the tumor-stroma ratio (TSR) in giant cell tumor of bone (GCTB). Hence, we aimed to investigate the distribution of TSR in GCTB and explore its correlation with various clinicopathologic factors, immune microenvironment, survival prognosis, and denosumab treatment responsiveness. METHODS: We conducted a multicenter cohort study comprising 426 GCTB patients treated at four centers. TSR was evaluated on hematoxylin and eosin-stained and immunofluorescent sections of tumor specimens. Immunohistochemistry was performed to assess CD3+, CD4+, CD8+, CD20+, PD-1+, PD-L1+, and FoxP3+ TIL subtypes as well as Ki-67 expression levels in 426 tissue specimens. These parameters were then analyzed for their correlations with patient outcomes [local recurrence-free survival (LRFS) and overall survival (OS)], clinicopathological features, and denosumab treatment responsiveness. RESULTS: Low TSR was significantly associated with poor LRFS and OS in both cohorts. Furthermore, TSR was also correlated with multiple clinicopathological features, TIL subtype expression, and denosumab treatment responsiveness. TSR demonstrated similar predictive capabilities as the conventional Campanacci staging system for predicting patients' LRFS and OS. CONCLUSION: The results of this study provide evidence supporting the use of TSR as a reliable prognostic tool in GCTB and as a predictor of denosumab treatment responsiveness. These findings may aid in developing individualized treatment strategies for GCTB patients in the future.

Aqueous Grown Quantum Dots with Robust Near-Infrared Fluorescence for Integrated Traumatic Brain Injury Diagnosis and Surgical Monitoring.

Surgical intervention is the most common first-line treatment for severe traumatic brain injuries (TBIs) associated with high intracranial pressure, while the complexity of these surgical procedures often results in complications. Surgeons often struggle to comprehensively evaluate the TBI status, making it difficult to select the optimal intervention strategy. Here, we introduce a fluorescence imaging-based technology that uses high-quality silver indium selenide-based quantum dots (QDs) for integrated TBI diagnosis and surgical guidance. These engineered, poly(ethylene glycol)-capped QDs emit in the near-infrared region, are resistant to phagocytosis, and importantly, are ultrastable after the epitaxial growth of an aluminum-doped zinc sulfide shell in the aqueous phase that renders the QDs resistant to long-term light irradiation and complex physiological environments. We found that intravenous injection of QDs enabled both the precise diagnosis of TBI in a mouse model and, more importantly, the comprehensive evaluation of the TBI status before, during, and after an operation to distinguish intracranial from superficial hemorrhages, provide real-time monitoring of the secondary hemorrhage, and guide the decision making on the evacuation of intracranial hematomas. This QD-based diagnostic and monitoring system could ultimately complement existing clinical tools for treating TBI, which may help surgeons improve patient outcomes and avoid unnecessary procedures.

Is Type 2 Diabetes Mellitus Associated with Spinal Degenerative Disorders?: Evidence from Observational and 2-Sample Mendelian Randomization Analyses.

BACKGROUND: Type 2 diabetes mellitus (T2DM) and spinal degenerative disorders (SDD) are common diseases that frequently coexist. However, both traditional observational studies and recent Mendelian randomization (MR) studies have demonstrated conflicting evidence on the association between T2DM and SDD. This comparative study explored and compared the association between T2DM and SDD using observational and MR analyses. METHODS: For observational analyses, cross-sectional studies (44,972 participants with T2DM and 403,095 participants without T2DM), case-control studies (38,234 participants with SDD and 409,833 participants without SDD), and prospective studies (35,550 participants with T2DM and 392,046 participants without T2DM with follow-up information until 2022) were performed to test the relationship between T2DM and SDD using individual-level data from the U.K. Biobank from 2006 to 2022. For MR analyses, the associations between single-nucleotide polymorphisms with SDD susceptibility obtained using participant data from the U.K. Biobank, which had 407,938 participants from 2006 to 2022, and the FinnGen Consortium, which had 227,388 participants from 2017 to 2022, and genetic predisposition to T2DM obtained using summary statistics from a pooled genome-wide association study involving 1,407,282 individuals were examined. The onset and severity of T2DM are not available in the databases being used. RESULTS: Participants with T2DM were more likely to have SDD than their counterparts. Logistic regression analysis identified T2DM as an independent risk factor for SDD, which was confirmed by the Cox proportional hazard model results. However, using single-nucleotide polymorphisms as instruments, the MR analyses demonstrated no causal relationship between T2DM and SDD. The lack of such an association was robust in the sensitivity analysis, and no pleiotropy was seen. CONCLUSIONS: Our results suggest that the association between T2DM and SDD may be method-dependent. Researchers and clinicians should be cautious in interpreting the association, especially the causal association, between T2DM and SDD. Our findings provide fresh insights into the association between T2DM and SDD by various analysis methods and guide future research and clinical efforts in the effective prevention and management of T2DM and SDD. LEVEL OF EVIDENCE: Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.

Complex immune microenvironment of chordoma: a road map for future treatment.

BACKGROUND: Chordoma, a rare bone tumor, presents limited treatment options and patients typically exhibit poor survival outcomes. While immunotherapy has shown promising results in treating various tumors, research on the immune microenvironment of chordomas is still in its early stages. Therefore, understanding how the immune microenvironment of chordomas influences the outcomes of immunotherapy is crucial. METHODS: We employed single-cell RNA sequencing (scRNA-seq), bulk RNA-seq, CellChat, gene set variation analysis, as well as calculation of immune features to further dissect the complex immune microenvironment of chordoma. RESULTS: Previous research by van Oost et al argued that compared with other sarcomas, chordomas typically exhibit an immunologically "hot" microenvironment, a conclusion with which we concur based on their research findings. Additionally, the authors suggest that T cell-mediated immunotherapy is feasible for the majority of chordomas. However, we are inclined to categorize them as an immune-excluded phenotype according to the latest classification methods, rather than persisting with the concepts of "cold" and "hot". Unlike them, we explored immune infiltration scores (IS), T lymphocyte scoring (TLS), and human leucocyte antigen class I (HLA-I) using Bulk RNA-seq data from 126 chordoma patients and found that higher IS, TLS, and higher HLA-I expression were associated with poorer patient prognosis. Additionally, CellChat analysis of scRNA-seq results from six chordoma patients revealed no direct interaction between T cells and tumor cells. CONCLUSIONS: These findings suggested that the efficacy of T cell-based immunotherapy may be limited or even ineffective for patients with chordoma.

Potential of digital chest radiography-based deep learning in screening and diagnosing pneumoconiosis: An observational study.

The diagnosis of pneumoconiosis is complex and subjective, leading to inevitable variability in readings. This is especially true for inexperienced doctors. To improve accuracy, a computer-assisted diagnosis system is used for more effective pneumoconiosis diagnoses. Three models (Resnet50, Resnet101, and DenseNet) were used for pneumoconiosis classification based on 1250 chest X-ray images. Three experienced and highly qualified physicians read the collected digital radiography images and classified them from category 0 to category III in a double-blinded manner. The results of the 3 physicians in agreement were considered the relative gold standards. Subsequently, 3 models were used to train and test these images and their performance was evaluated using multi-class classification metrics. We used kappa values and accuracy to evaluate the consistency and reliability of the optimal model with clinical typing. The results showed that ResNet101 was the optimal model among the 3 convolutional neural networks. The AUC of ResNet101 was 1.0, 0.9, 0.89, and 0.94 for detecting pneumoconiosis categories 0, I, II, and III, respectively. The micro-average and macro-average mean AUC values were 0.93 and 0.94, respectively. The accuracy and Kappa values of ResNet101 were 0.72 and 0.7111 for quadruple classification and 0.98 and 0.955 for dichotomous classification, respectively, compared with the relative standard classification of the clinic. This study develops a deep learning based model for screening and staging of pneumoconiosis is using chest radiographs. The ResNet101 model performed relatively better in classifying pneumoconiosis than radiologists. The dichotomous classification displayed outstanding performance, thereby indicating the feasibility of deep learning techniques in pneumoconiosis screening.

Single-cell RNA sequencing reveals the MIF/ACKR3 receptor-ligand interaction between neutrophils and nucleus pulposus cells in intervertebral disc degeneration.

OBJECTIVES: To unravel the heterogeneity and function of microenvironmental neutrophils during intervertebral disc degeneration (IDD). METHODS: Single-cell RNA sequencing (scRNA-seq) was utilized to dissect the cellular landscape of neutrophils in intervertebral disc (IVD) tissues and their crosstalk with nucleus pulposus cells (NPCs). The expression levels of macrophage migration inhibitory factor (MIF) and ACKR3 in IVD tissues were detected. The MIF/ACKR3 axis was identified and its effects on IDD were investigated in vitro and in vivo. RESULTS: We sequenced here 71520 single cells from 5 control and 9 degenerated IVD samples using scRNA-seq. We identified a unique cluster of neutrophils abundant in degenerated IVD tissues that highly expressed MIF and was functionally enriched in extracellular matrix organization (ECMO). Cell-to-cell communication analyses showed that this ECMO-neutrophil subpopulation was closely interacted with an effector NPCs subtype, which displayed high expression of ACKR3. Further analyses revealed that MIF was positively correlated with ACKR3 and functioned via directly binding to ACKR3 on effector NPCs. MIF inhibition attenuated degenerative changes of NPCs and extracellular matrix, which could be partially reversed by ACKR3 overexpression. Clinically, a significant correlation of high MIF/ACKR3 expression with advanced IDD grade was observed. Furthermore, we also found a positive association between MIF+ ECMO-neutrophil counts and ACKR3+ effector NPCs density as well as higher expression of the MIF/ACKR3 signaling in areas where these two cell types were neighbors. CONCLUSIONS: These data suggest that ECMO-neutrophil promotes IDD progression by their communication with NPCs via the MIF/ACKR3 axis, which may shed light on therapeutic strategies.

Comparative safety of monoclonal antibodies in chronic inflammatory airway diseases (chronic sinusitis with nasal polyposis and asthma): A network meta-analysis.

OBJECTIVE: Several monoclonal antibodies (MoAbs) targeting specific type 2 immune reactions have been developed as innovative therapeutic approaches for chronic inflammatory airway diseases, such as chronic sinusitis with nasal polyps (CRSwNP) and asthma. However, the clinical safety of these MoAbs and how to choose them are not clear. Therefore, we aimed to assess the systemic drug- and dose-based safety of MoAbs in chronic airway inflammation using network meta-analysis (NMA). METHODS: Electronic databases were systematically searched for relevant studies published in English between January 2009 and December 2022. Eligible studies must have clearly reported adverse events (AEs) among the MoAbs' safety data. RESULTS: 1). Regarding serious AEs, mepolizumab was significantly safer than placebo; in terms of permanent treatment discontinuation, reslizumab and dupilumab were significantly safer than benralizumab. 2). Regarding asthma worsening, dupilumab was associated with the best safety profile; was safer than dupilumab/300 mg/q2-4w. 3). In terms of injection-site reactions, dupilumab posed a higher risk than placebo; dupilumab/300 mg/qw posed a higher risk than dupilumab/300 mg/q2w and dupilumab/300 mg/q2-4w; lebrikizumab/250 mg/q4w posed a higher risk than lebrikizumab/37.5 mg/q4w; mepolizumab/100 mg/q4w posed a higher risk than mepolizumab/75 mg/q4w; benralizumab/30 mg/q4-8w posed a higher risk than benralizumab/20 mg/q4-8w. 4) In CRSwNP patients combined with asthma, the risks of experiencing AEs were not increased. CONCLUSION: Overall, biologics are safe and well tolerated in chronic inflammatory airway disease. This drug- and dose-based NMA provides further evidence on the different safety profiles of different emerging MoAbs. This information may help guide rational drug use and provide clinical recommendations for choosing MoAbs. TRIAL REGISTRATION: SYSTEMATIC REVIEW REGISTRATION (PROSPERO #CRD42023387610).

Glycyrol Relieves Ulcerative Colitis by Promoting the Fusion of ZO-1 with the Cell Membrane through the Enteric Glial Cells GDNF/RET Pathway.

The damage to the mechanical barrier of the intestinal mucosa is the initiating factor and the core link of the progression of ulcerative colitis (UC). Protecting the mechanical barrier of the intestinal mucosa is of great significance for improving the health status of UC patients. ZO-1 is a key scaffold protein of the mechanical barrier of the intestinal mucosa, and its fusion with the membrane of the intestinal epithelium is a necessary condition to maintain the integrity of the mechanical barrier of the intestinal mucosa. Enteric glial cells (EGCs) play an important role in the maintenance of intestinal homeostasis and have become a new target for regulating intestinal health in recent years. In this study, we found that glycyrol (GC), a representative coumarin compound isolated from Licorice (Glycyrrhiza uralensis Fisch, used for medicine and food), can alleviate UC by promoting the production of neurotrophic factor GDNF in mice EGCs. Specifically, we demonstrated that GC promotes the production of GDNF, then activates its receptor RET, promotes ZO-1 fusion with cell membranes, and protects the intestinal mucosal mechanical barrier. The results of this study can provide new ideas for the prevention and treatment of UC.

Neddylation activated TRIM25 desensitizes triple-negative breast cancer to paclitaxel via TFEB-mediated autophagy.

BACKGROUND: Paclitaxel (PTX) treatment resistance is an important factor leading to poor prognosis in triple-negative breast cancer (TNBC), therefore there is an urgent need to identify new target for combination therapy. Neddylation is a post-translational process that introduces a ubiquitin-like protein called neural precursor cell expressed developmentally downregulated protein 8 (NEDD8). Previous studies have found that neddylation is activated in multiple tumors, but its relationship with PTX chemotherapy sensitivity has not been reported. METHODS: Differences in UBC12 and NEDD8 expression levels between PTX-sensitive and PTX-insensitive TNBC tissues were validated using public databases and immunohistochemistry. The in vitro and in vivo functional experiments were used to observe the effect of neddylation inhibition combined with PTX therapy on tumor progression. Co-IP, western blot and PCR assays were used to investigate the molecular mechanisms. Molecular docking was used to simulate the protein binding of UBC12 and TRIM25. Molecular dynamics simulation was used to observe the changes in TRIM25 protein conformation. RESULTS: We found that in TNBC that is insensitive to PTX, NEDD8 and NEDD8 conjugating enzyme UBC12 are highly expressed. Treatment with the NEDD8-activating enzyme (NAE) inhibitor mln4924 or knockdown of UBC12 significantly increased the sensitivity of the tumor to PTX, and this increase in sensitivity is related to UBC12-mediated autophagy activation. Mechanistically, UBC12 can transfer NEDD8 to E3 ubiquitin ligase tripartite motif containing 25 (TRIM25) at K117. Molecular dynamics simulations indicate that the neddylation modification of TRIM25 reduces steric hindrance in its RING domain, facilitating the binding of TRIM25 and ubiquitylated substrates. Subsequently, TRIM25 promotes the nuclear translocation of transcription factor EB (TFEB) and transcription of autophagy related genes by increasing K63-polyubiquitination of TFEB, thereby reducing tumor sensitivity to PTX. CONCLUSIONS: Neddylation is activated in PTX-insensitive TNBC. Specifically, autophagy gene transcriptional activation mediated by the UBC12/TRIM25/TFEB axis reduces TNBC sensitivity to PTX. Neddylation suppression combination with PTX treatment shows a synergistic anti-tumor effect.

Targeted delivery of type I TGF-ß receptor-mimicking peptide to fibrotic kidney for improving kidney fibrosis therapy via enhancing the inhibition of TGF-ß1/Smad and p38 MAPK pathways.

Renal fibrosis is a representative pathological feature of various chronic kidney diseases, and efficient treatment is needed. Interstitial myofibroblasts are a key driver of kidney fibrosis, which is dependent on the binding of TGF-ß1 to type I TGF-ß receptor (TßRI) and TGF-ß1-related signaling pathways. Therefore, attenuating TGF-ß1 activity by competing with TGF-ß1 in myofibroblasts is an ideal strategy for treating kidney fibrosis. Recently, a novel TßRI-mimicking peptide RIPΔ demonstrated a high affinity for TGF-ß1. Thus, it could be speculated that RIPΔ may be used for anti-fibrosis therapy. Platelet-derived growth factor ß receptor (PDGFßR) is highly expressed in fibrotic kidney. In this study, we found that target peptide Z-RIPΔ, which is RIPΔ modified with PDGFßR-specific affibody ZPDGFßR, was specifically and highly taken up by TGF-ß1-activated NIH3T3 fibroblasts. Moreover, Z-RIPΔ effectively inhibited the myofibroblast proliferation, migration and fibrosis response in vitro. In vivo and ex vivo experiments showed that Z-RIPΔ specifically targeted fibrotic kidney, improved the damaged renal function, and ameliorated kidney histopathology and renal fibrosis in UUO mice. Mechanistic studies showed that Z-RIPΔ hold the stronger inhibition of the TGF-ß1/Smad and TGF-ß1/p38 pathways than unmodified RIPΔ in vitro and in vivo. Furthermore, systemic administration of Z-RIPΔ to UUO mice led to minimal toxicity to major organs. Taken together, RIPΔ modified with ZPDGFßR increased its therapeutic efficacy and reduced its systemic toxicity, making it a potential candidate for targeted therapy for kidney fibrosis.

Development of a predictive nomogram for in-hospital death risk in multimorbid patients with hepatocellular carcinoma undergoing Palliative Locoregional Therapy.

Patients diagnosed with hepatocellular carcinoma (HCC) often present with multimorbidity, significantly contributing to adverse outcomes, particularly in-hospital mortality. This study aimed to develop a predictive nomogram to assess the impact of comorbidities on in-hospital mortality risk in HCC patients undergoing palliative locoregional therapy. We retrospectively analyzed data from 345 hospitalized HCC patients who underwent palliative locoregional therapy between January 2015 and December 2022. The nomogram was constructed using independent risk factors such as length of stay (LOS), hepatitis B virus (HBV) infection, hypertension, chronic obstructive pulmonary disease (COPD), anemia, thrombocytopenia, liver cirrhosis, hepatic encephalopathy (HE), N stage, and microvascular invasion. The model demonstrated high predictive accuracy with an AUC of 0.908 (95% CI: 0.859-0.956) for the overall dataset, 0.926 (95% CI: 0.883-0.968) for the training set, and 0.862 (95% CI: 0.728-0.994) for the validation set. Calibration curves indicated a strong correlation between predicted and observed outcomes, validated by statistical tests. Decision curve analysis (DCA) and clinical impact curves (CIC) confirmed the model's clinical utility in predicting in-hospital mortality. This nomogram offers a practical tool for personalized risk assessment in HCC patients undergoing palliative locoregional therapy, facilitating informed clinical decision-making and improving patient management.

Eigendecomposition-free inverse design of meta-optics devices.

The inverse design of meta-optics has received much attention in recent years. In this paper, we propose a GPU-friendly inverse design framework based on improved eigendecomposition-free rigorous diffraction interface theory, which offers up to 16.2 × speedup over the traditional inverse design based on rigorous coupled-wave analysis. We further improve the framework's flexibility by introducing a hybrid parameterization combining neural-implicit and traditional shape optimization. We demonstrate the effectiveness of our framework through intricate tasks, including the inverse design of reconfigurable free-form meta-atoms.

Can low-carbon cities pilot policy promote enterprise sustainable development? Quasi-experimental evidence from China.

With the predicament of sustainable improvement in traditional cities, the low-carbon city pilot policy (LCCPP), as a novel development mode, provides thinking for resolving the tensions of green development, resource conservation and environmental protection among firms. Using Chinese A-share listed companies panel data during 2007-2019, this study adopts the difference-in-differences model to explore the impact of LCCPP on firm green innovation. Based on theoretical analysis, LCCPP-driven environmental rules have the impact of encouraging business green innovation. The relationship between LCCPP and green innovation is strengthened by external media attention and organizational redundancy resources. The mechanism study shows that the incentive effect of LCCPP on firm green innovation is mainly due to the improvement of enterprises' green total factor productivity and financial stability. In addition, the heterogeneity analysis shows that the LCCPP has significantly positive effects in promoting green innovation in high-carbon industries and state-owned enterprises. This research contributes to the understanding of city-level low-carbon policies as a driving force for corporate green innovation, offering practical implications for policymakers and businesses striving for sustainability.

Automatic segmentation of 15 critical anatomical labels and measurements of cardiac axis and cardiothoracic ratio in fetal four chambers using nnU-NetV2.

BACKGROUND: Accurate segmentation of critical anatomical structures in fetal four-chamber view images is essential for the early detection of congenital heart defects. Current prenatal screening methods rely on manual measurements, which are time-consuming and prone to inter-observer variability. This study develops an AI-based model using the state-of-the-art nnU-NetV2 architecture for automatic segmentation and measurement of key anatomical structures in fetal four-chamber view images. METHODS: A dataset, consisting of 1,083 high-quality fetal four-chamber view images, was annotated with 15 critical anatomical labels and divided into training/validation (867 images) and test (216 images) sets. An AI-based model using the nnU-NetV2 architecture was trained on the annotated images and evaluated using the mean Dice coefficient (mDice) and mean intersection over union (mIoU) metrics. The model's performance in automatically computing the cardiac axis (CAx) and cardiothoracic ratio (CTR) was compared with measurements from sonographers with varying levels of experience. RESULTS: The AI-based model achieved a mDice coefficient of 87.11% and an mIoU of 77.68% for the segmentation of critical anatomical structures. The model's automated CAx and CTR measurements showed strong agreement with those of experienced sonographers, with respective intraclass correlation coefficients (ICCs) of 0.83 and 0.81. Bland-Altman analysis further confirmed the high agreement between the model and experienced sonographers. CONCLUSION: We developed an AI-based model using the nnU-NetV2 architecture for accurate segmentation and automated measurement of critical anatomical structures in fetal four-chamber view images. Our model demonstrated high segmentation accuracy and strong agreement with experienced sonographers in computing clinically relevant parameters. This approach has the potential to improve the efficiency and reliability of prenatal cardiac screening, ultimately contributing to the early detection of congenital heart defects.

A comparative study of progressive failure of granite and marble rock bridges under direct shearing.

Shear failure of rock bridges is an important process in geological phenomena, including landslides and earthquakes. However, the progressive failure of natural rock bridges has not yet been fully understood. In this work, we carried out direct shearing experiments on both granite and marble rock bridges, and applied acoustic emission (AE) monitoring throughout the experiments. With the mechanical curves and the evolution of AE activity (including AE energy rate and b value), the failure of rock bridges can be divided into three pre-failure phases and one ultimate failure phases. We analyzed the effects of normal stress and lithology on the pre-failure phases. We noted that with the increasing of normal stress, the length of stable cracking phase decreases and the length of unstable cracking phase slightly increases, except for marble rock bridges at high normal stress, which maintains a great proportion of stable cracking phase that possibly results from the great off-fault damage. Increasing normal stress also suppresses the dilation of granite rock bridges, but has a different effect on marble rock bridges, which also suggests the effect of lithology on failure modes.

GBDT Method Integrating Feature-Enhancement and Active-Learning Strategies-Sea Ice Thickness Inversion in Beaufort Sea.

Sea ice, as an important component of the Earth's ecosystem, has a profound impact on global climate and human activities due to its thickness. Therefore, the inversion of sea ice thickness has important research significance. Due to environmental and equipment-related limitations, the number of samples available for remote sensing inversion is currently insufficient. At high spatial resolutions, remote sensing data contain limited information and noise interference, which seriously affect the accuracy of sea ice thickness inversion. In response to the above issues, we conducted experiments using ice draft data from the Beaufort Sea and designed an improved GBDT method that integrates feature-enhancement and active-learning strategies (IFEAL-GBDT). In this method, the incident angle and time series are used to perform spatiotemporal correction of the data, reducing both temporal and spatial impacts. Meanwhile, based on the original polarization information, effective multi-attribute features are generated to expand the information content and improve the separability of sea ice with different thicknesses. Taking into account the growth cycle and age of sea ice, attributes were added for month and seawater temperature. In addition, we studied an active learning strategy based on the maximum standard deviation to select more informative and representative samples and improve the model's generalization ability. The improved GBDT model was used for training and prediction, offering advantages in dealing with nonlinear, high-dimensional data, and data noise problems, further expanding the effectiveness of feature-enhancement and active-learning strategies. Compared with other methods, the method proposed in this paper achieves the best inversion accuracy, with an average absolute error of 8 cm and a root mean square error of 13.7 cm for IFEAL-GBDT and a correlation coefficient of 0.912. This research proves the effectiveness of our method, which is suitable for the high-precision inversion of sea ice thickness determined using Sentinel-1 data.