Regorafenib with immunotherapy versus regorafenib alone as second-line treatment for hepatocellular carcinoma: A multicenter real-world study.

INTRODUCTION: Regorafenib remains the standard and widely used second-line strategy for advanced hepatocellular carcinoma (HCC). There is still a lack of large-scale multicenter real-world evidence concerning the concurrent use of regorafenib with immune checkpoint inhibitors (ICI). This study aims to evaluate whether combining regorafenib with ICI provides greater clinical benefit than regorafenib monotherapy as second-line therapy for advanced HCC under real-world circumstances. PATIENTS AND METHODS: The study included 208 patients from five medical facilities. One hundred forty-three patients received regorafenib plus ICI combination therapy, while 65 patients received regorafenib monotherapy. Propensity score matching (PSM) analysis was employed. RESULTS: The regorafenib plus ICI group demonstrated significantly higher objective response rate (24.3% vs. 10.3%, after PSM, p = 0.030) and disease control rate (79.4% vs. 50.0%, after PSM, p < 0.001) compared to the regorafenib monotherapy group based on mRECIST criteria. Median progression-free survival (7.9 vs. 3.2 months, after PSM, p < 0.001) and overall survival (25.6 vs. 16.4 months, p = 0.010, after PSM) were also considerably longer in the regorafenib plus ICI group. The incidence of Grades 3-4 treatment-related adverse events (TRAEs) was marginally greater in the regorafenib plus ICI group than in the regorafenib group (23.8% vs. 20.0%, p = 0.546). Notably, there were no instances of treatment-related mortality or emergence of new TRAEs in any treatment group. CONCLUSION: The combination of regorafenib and ICI shows potential as a viable second-line treatment for advanced HCC, exhibiting favorable efficacy while maintaining a tolerable safety profile in contrast to regorafenib monotherapy.

Elastic Properties of Low-Dimensional Single-Crystalline Dielectric Oxides through Controlled Large-Area Wrinkle Generation.

Freestanding single-crystalline SrTiO3 membranes, as high-κ dielectrics, hold significant promise as the gate dielectric in two-dimensional (2D) flexible electronics. Nevertheless, the mechanical properties of the SrTiO3 membranes, such as elasticity, remain a critical piece of the puzzle to adequately address the viability of their applications in flexible devices. Here, we report statistical analysis on plane-strain effective Young's modulus of large-area SrTiO3 membranes (5 × 5 mm2) over a series of thicknesses (from 6.5 to 32.2 nm), taking advantage of a highly efficient buckling-based method, which reveals its evident thickness-dependent behavior ranging from 46.01 to 227.17 GPa. Based on microscopic and theoretical results, we elucidate these thickness-dependent behaviors and statistical data deviation with a bilayer model, which consists of a surface layer and a bulk-like layer. The analytical results show that the ∼3.1 nm surface layer has a significant elastic softening compared to the bulk-like layer, while the extracted modulus of the bulk-like layer shows a variation of ∼40 GPa. This variation is considered as a combined contribution from oxygen deficiency presenting in SrTiO3 membranes, and the alignment between applied strain and the crystal orientation. Upon comparison of the extracted elastic properties and electrostatic control capability to those of other typical gate dielectrics, the superior performance of single-crystalline SrTiO3 membranes has been revealed in the context of flexible gate dielectrics, indicating the significant potential of their application in high-performance flexible 2D electronics.

[Impact of anemia on incidence of perioperative lower limb deep vein thrombosis in patients undergoing total hip arthroplasty].

Objective: To explore the impact of anemia on the incidence of perioperative lower limb deep vein thrombosis (DVT) in patients undergoing total hip arthroplasty (THA). Methods: A retrospective analysis was conducted on clinical data of 1 916 non-fracture patients who underwent THA between September 2015 and December 2021, meeting the selection criteria. Among them, there were 811 male and 1 105 female patients, aged between 18 and 94 years with an average of 59.2 years. Among the patients, 213 were diagnosed with anemia, while 1 703 were not. Preoperative DVT was observed in 55 patients, while 1 861 patients did not have DVT preoperatively (of which 75 patients developed new-onset DVT postoperatively). Univariate analysis was performed on variables including age, gender, body mass index (BMI), diabetes, hypertension, history of tumors, history of thrombosis, history of smoking, revision surgery, preoperative D-dimer positivity (≥0.5 mg/L), presence of anemia, operation time, intraoperative blood loss, transfusion requirement, and pre- and post-operative levels of red blood cells, hemoglobin, hematocrit, and platelets. Furthermore, logistic regression was utilized for multivariate analysis to identify risk factors associated with DVT formation. Results: Univariate analysis showed that age, gender, hypertension, revision surgery, preoperative levels of red blood cells, preoperative hemoglobin, preoperative D-dimer positivity, and anemia were influencing factors for preoperative DVT ( P<0.05). Further logistic regression analysis indicated that age (>60 years old), female, preoperative D-dimer positivity, and anemia were risk factors for preoperative DVT ( P<0.05). Univariate analysis also revealed that age, female, revision surgery, preoperative D-dimer positivity, anemia, transfusion requirement, postoperative level of red blood cells, and postoperative hemoglobin level were influencing factors for postoperative new-onset DVT ( P<0.05). Further logistic regression analysis indicated that age (>60 years old), female, and revision surgery were risk factors for postoperative new-onset DVT ( P<0.05). Conclusion: The incidence of anemia is higher among patients with preoperative DVT for THA, and anemia is an independent risk factor for preoperative DVT occurrence in THA. While anemia may not be an independent risk factor for THA postoperative new-onset DVT, the incidence of anemia is higher among patients with postoperative new-onset DVT.

Association of AST/ALT ratio with 90-day outcomes in patients with acute exacerbation of chronic liver disease: a prospective multicenter cohort study in China.

Background and aim: A high aspartate aminotransferase/alanine aminotransferase (AST/ALT) ratio is associated with liver injury in liver disease; however, no data exist regarding its relationship with 90-day prognosis in patients with acute exacerbation of chronic liver disease. Methods: In this study, 3,758 participants (955 with advanced fibrosis and 2,803 with cirrhosis) from the CATCH-LIFE cohort in China were included. The relationships between different AST/ALT ratios and the risk of adverse 90-day outcomes (death or liver transplantation) were determined in patients with cirrhosis or hepatitis B virus (HBV)-associated advanced fibrosis, respectively. Results: In the patients with HBV-associated advanced fibrosis, the risk of 90-day adverse outcomes increased with AST/ALT ratio; after adjusting for all confounding factors, the risk of adverse 90-day outcomes was the highest when AST/ALT ratio was more than 1.08 (OR = 6.91 [95% CI = 1.789-26.721], p = 0.005), and the AST/ALT ratio of >1.9 accelerated the development of adverse outcomes. In patients with cirrhosis, an AST/ALT ratio > 1.38 increased the risk of adverse 90-day outcomes in all univariables (OR = 1.551 [95% CI = 1.216-1.983], p < 0.001) and multivariable-adjusted analyses (OR = 1.847 [95% CI = 1.361-2.514], p < 0.001), and an elevated AST/ALT ratio (<2.65) accelerated the incidence of 90-day adverse outcomes. An AST/ALT ratio of >1.38 corresponded with a more than 20% incidence of adverse outcomes in patients with cirrhosis. Conclusion: The AST/ALT ratio is an independent risk factor for adverse 90-day outcomes in patients with cirrhosis and HBV-associated advanced fibrosis. The cutoff values of the AST/ALT ratio could help clinicians monitor the condition of patients when making clinical decisions.

Domino Michael/Oxa-Michael Reactions of the Unsymmetric Double Michael Acceptor for Access to Bicyclic Furo[2,3-b]pyrrole.

By creating an unsymmetric double Michael acceptor 1, we were able to synthesize the nonaromatic-fused bicyclic furo[2,3-b]pyrrole nucleus using a domino Michael/oxa-Michael reaction. Adopting benzoyl acetonitrile 2d (CN as the electron-withdrawing group) as a substrate, we discovered a (DHQ)2AQN-catalyzed method for high diastereo- and enantioselectivity of those products. The reaction path has been determined by isolating the reaction intermediates, and density functional theory calculations support these findings. Beyond providing a synthetic approach, this work illustrated the compounds' possible use in antitumor activity.

Characterization and prevention of flame burns caused by electric bicycle battery chargers: results from a 7-year experience.

OBJECTIVE: With the recent exponential increase in express deliveries across China, the number of patients with flame burns caused by electric bicycle battery chargers (BEBBC) has markedly increased in burn units. In this study, we aimed to characterize BEBBC to systematically explore measures to prevent their occurrence. METHODS: We performed a retrospective chart review of patients with flame burns who visited the Burn Department of Rui Jin Hospital between January 2015 and December 2021. RESULTS: Sixty-three patients with BEBBC and 1412 with types of other flame burn were included in this study. Fifty-six of the 63 BEBBC cases occurred between 9 pm and 7 am. BEBBC incidents involved a higher incidence of group burn in which multiple individuals were affected. Non-local patients with BEBBC were significantly younger than their local counterparts. BEBBC had a higher mortality than types of other flame burn. CONCLUSIONS: The rising incidence of BEBBC calls for greater attention because of the associated high mortality and heavy burden on society. Enacting related legislation, disseminating information to the public, and improving treatment to control infection can help prevent BEBBC, increase its cure rate, and reduce patient mortality.

MazF-rolling circle amplification combined MALDI-TOF MS for site-specific detection of N6-methyladenosine RNA.

N6-methyladenosine (m6A) is one of the most abundant chemical modifications in RNA and has vital significance in cellular processes and tumor development. However, the accurate analysis of site-specific m6A modification remains a challenge. In this work, a MazF endoribonuclease activated rolling circle amplification (MazF-RCA) combined MALDI-TOF MS assay is developed for the detection of site-specific m6A-RNA. MazF endoribonuclease can specifically cleave the ACA motif, leaving methylated (m6A)CA motif intact. The intact methylated RNA can then be amplified through rolling circle amplification, and the generated reporter oligonucleotides are detected by MALDI-TOF MS. The assay exhibits good quantification ability, presenting a wide linear range (100 fM to 10 nM) with the limit-of-detection lower than 100 fM. Additionally, the assay can accurately detect methylated RNA in the presence of large amount of non-methylated RNA with a relative abundance of methylated RNA down to 0.5%. The developed assay was further applied to detect m6A-RNA spiked in MCF-7 cell RNA extracts, with the recovery rates in the range of 90.64-106.93%. The present assay provides a novel platform for the analysis of site-specific m6A-RNA at high specificity and sensitivity, which can promote the study of RNA methylation in clinical and biomedical research.

Facet-Dependent Oxygen Evolution Reaction Activity of IrO2 from Quantum Mechanics and Experiments.

The diversity of chemical environments present on unique crystallographic facets can drive dramatic differences in catalytic activity and the reaction mechanism. By coupling experimental investigations of five different IrO2 facets and theory, we characterize the detailed elemental steps of the surface redox processes and the rate-limiting processes for the oxygen evolution reaction (OER). The predicted complex evolution of surface adsorbates and the associated charge transfer as a function of applied potential matches well with the distinct redox features observed experimentally for the five facets. Our microkinetic model from grand canonical quantum mechanics (GC-QM) calculations demonstrates mechanistic differences between nucleophilic attack and O-O coupling across facets, providing the rates as a function of applied potential. These GC-QM calculations explain the higher OER activity observed on the (100), (001), and (110) facets and the lower activity observed for the (101) and (111) facets. This combined study with theory and experiment brings new insights into the structural features that either promote or hinder the OER activity of IrO2, which are expected to provide parallels in structural effects on other oxide surfaces.

Integrated genome-wide association and transcriptomic analysis to identify receptor kinase genes to stripe rust resistance in wheat germplasm from southwestern China.

Stripe rust of wheat, caused by Puccinia striiformis f. sp. tritici (Pst), is one of the most important diseases of wheat worldwide. Identification of new and elite Pst-resistance loci or genes has the potential to enhance overall resistance to this pathogen. Here, we conducted an integrated genome-wide association study (GWAS) and transcriptomic analysis to screen for loci associated with resistance to stripe rust in 335 accessions from Yunnan, including 311 landraces and 24 cultivars. Based on the environmental phenotype, we identified 113 protein kinases significantly associated with Pst resistance using mixed linear model (MLM) and generalized linear model (GLM) models. Transcriptomic analysis revealed that 52 of 113 protein kinases identified by GWAS were up and down regulated in response to Pst infection. Among these genes, a total of 15 receptor kinase genes were identified associated with Pst resistance. 11 candidate genes were newly discovered in Yunnan wheat germplasm. Our results revealed that resistance alleles to stripe rust were accumulated in Yunnan wheat germplasm, implying direct or indirect selection for improving stripe rust resistance in elite wheat breeding programs.

Robust Superconductivity in Infinite-Layer Nickelates.

The recent discovery of nickelate superconductivity represents an important step toward understanding the four-decade mastery of unconventional high-temperature superconductivity. However, the synthesis of the infinite-layer nickelate superconductors shows great challenges. Particularly, surface capping layers are usually unitized to facilitate the sample synthesis. This leads to an important question whether nickelate superconductors with d9 configuration and ultralow valence of Ni1+ are in metastable state and whether nickelate superconductivity can be robust? In this work, a series of redox cycling experiments are performed across the phase transition between perovskite Nd0.8Sr0.2NiO3 and infinite-layer Nd0.8Sr0.2NiO2. The infinite-layer Nd0.8Sr0.2NiO2 is quite robust in the redox environment and can survive the cycling experiments with unchanged crystallographic quality. However, as the cycling number goes on, the perovskite Nd0.8Sr0.2NiO3 shows structural degradation, suggesting stability of nickelate superconductivity is not restricted by the ultralow valence of Ni1+, but by the quality of its perovskite precursor. The observed robustness of infinite-layer Nd0.8Sr0.2NiO2 up to ten redox cycles further indicates that if an ideal high-quality perovskite precursor can be obtained, infinite-layer nickelate superconductivity can be very stable and sustainable under environmental conditions. This work provides important implications for potential device applications for nickelate superconductors.

Constructing copper Phthalocyanine/Molybdenum disulfide (CuPc/MoS2) S-scheme heterojunction with S-rich vacancies for enhanced Visible-Light photocatalytic CO2 reduction.

Constructing a heterojunction by combining two semiconductors with similar band structures is a successful approach to obtaining photocatalysts with high efficiency. Herein, a CuPc/DR-MoS2 heterojunction involving copper phthalocyanine (CuPc) and molybdenum disulfide with S-rich vacancies (13.66%) is successfully prepared by the facile hydrothermal method. Experimental results and theoretical calculations firmly demonstrated that photoelectrons exhibit an S-scheme charge transfer mechanism in the CuPc/DR-MoS2 heterojunction. The S-scheme heterojunction system has proven significant advantages in promoting the charge separation and transfer of photogenerated carriers, enhancing visible-light responsiveness, and achieving robust photoredox capability. As a result, the optimized 3CuPc/DR-MoS2 S-scheme heterojunction exhibits photocatalytic yields of CO and CH4 at 200 and 111.6 µmol g-1h-1, respectively. These values are four times and 4.5 times greater than the photocatalytic yields of pure DR-MoS2. This study offers novel perspectives on the advancement of innovative and highly effective heterojunction photocatalysts.

Applying image features of proximal paracancerous tissues in predicting prognosis of patients with hepatocellular carcinoma.

BACKGROUND: Most of the methods using digital pathological image for predicting Hepatocellular carcinoma (HCC) prognosis have not considered paracancerous tissue microenvironment (PTME), which are potentially important for tumour initiation and metastasis. This study aimed to identify roles of image features of PTME in predicting prognosis and tumour recurrence of HCC patients. METHODS: We collected whole slide images (WSIs) of 146 HCC patients from Sun Yat-sen Memorial Hospital (SYSM dataset). For each WSI, five types of regions of interests (ROIs) in PTME and tumours were manually annotated. These ROIs were used to construct a Lasso Cox survival model for predicting the prognosis of HCC patients. To make the model broadly useful, we established a deep learning method to automatically segment WSIs, and further used it to construct a prognosis prediction model. This model was tested by the samples of 225 HCC patients from the Cancer Genome Atlas Liver Hepatocellular Carcinoma (TCGA-LIHC). RESULTS: In predicting prognosis of the HCC patients, using the image features of manually annotated ROIs in PTME achieved C-index 0.668 in the SYSM testing dataset, which is higher than the C-index 0.648 reached by the model only using image features of tumours. Integrating ROIs of PTME and tumours achieved C-index 0.693 in the SYSM testing dataset. The model using automatically segmented ROIs of PTME and tumours achieved C-index of 0.665 (95% CI: 0.556-0.774) in the TCGA-LIHC samples, which is better than the widely used methods, WSISA (0.567), DeepGraphSurv (0.593), and SeTranSurv (0.642). Finally, we found the Texture SumAverage Skew HV on immune cell infiltration and Texture related features on desmoplastic reaction are the most important features of PTME in predicting HCC prognosis. We additionally used the model in prediction HCC recurrence for patients from SYSM-training, SYSM-testing, and TCGA-LIHC datasets, indicating the important roles of PTME in the prediction. CONCLUSIONS: Our results indicate image features of PTME is critical for improving the prognosis prediction of HCC. Moreover, the image features related with immune cell infiltration and desmoplastic reaction of PTME are the most important factors associated with prognosis of HCC.

Importance of soil moisture conservation in mitigating climate change.

A troubling feedback loop, where drier soil contributes to hotter climates, has been widely recognized. This study, drawing on climate model simulations, reveals that maintaining current global soil moisture levels could significantly alleviate 32.9% of land warming under low-emission scenarios. This action could also postpone reaching critical warming thresholds of 1.5 °C and 2.0 °C by at least a decade. Crucially, preserving soil moisture at current levels could prevent noticeable climate change impacts across 42% of the Earth's land, a stark deviation from projections suggesting widespread impacts before the 2060s. To combat soil drying, afforestation in mid-to-low latitude regions within the next three decades is proposed as an effective strategy to increase surface water availability. This underscores the substantial potential of nature-based solutions for managing soil moisture, benefiting both climate change mitigation and ecological enhancement.

Contigs directed gene annotation (ConDiGA) for accurate protein sequence database construction in metaproteomics.

BACKGROUND: Microbiota are closely associated with human health and disease. Metaproteomics can provide a direct means to identify microbial proteins in microbiota for compositional and functional characterization. However, in-depth and accurate metaproteomics is still limited due to the extreme complexity and high diversity of microbiota samples. It is generally recommended to use metagenomic data from the same samples to construct the protein sequence database for metaproteomic data analysis. Although different metagenomics-based database construction strategies have been developed, an optimization of gene taxonomic annotation has not been reported, which, however, is extremely important for accurate metaproteomic analysis. RESULTS: Herein, we proposed an accurate taxonomic annotation pipeline for genes from metagenomic data, namely contigs directed gene annotation (ConDiGA), and used the method to build a protein sequence database for metaproteomic analysis. We compared our pipeline (ConDiGA or MD3) with two other popular annotation pipelines (MD1 and MD2). In MD1, genes were directly annotated against the whole bacterial genome database; in MD2, contigs were annotated against the whole bacterial genome database and the taxonomic information of contigs was assigned to the genes; in MD3, the most confident species from the contigs annotation results were taken as reference to annotate genes. Annotation tools, including BLAST, Kaiju, and Kraken2, were compared. Based on a synthetic microbial community of 12 species, it was found that Kaiju with the MD3 pipeline outperformed the others in the construction of protein sequence database from metagenomic data. Similar performance was also observed with a fecal sample, as well as in silico mixed datasets of the simulated microbial community and the fecal sample. CONCLUSIONS: Overall, we developed an optimized pipeline for gene taxonomic annotation to construct protein sequence databases. Our study can tackle the current taxonomic annotation reliability problem in metagenomics-derived protein sequence database and can promote the in-depth metaproteomic analysis of microbiome. The unique metagenomic and metaproteomic datasets of the 12 bacterial species are publicly available as a standard benchmarking sample for evaluating various analysis pipelines. The code of ConDiGA is open access at GitHub for the analysis of microbiota samples. Video Abstract.

Heterogeneity of hepatocellular carcinoma: from mechanisms to clinical implications.

Hepatocellular Carcinoma (HCC) is one of the most common types of primary liver cancer. Current treatment options have limited efficacy against this malignancy, primarily owing to difficulties in early detection and the inherent resistance to existing drugs. Tumor heterogeneity is a pivotal factor contributing significantly to treatment resistance and recurrent manifestations of HCC. Intratumoral heterogeneity is an important aspect of the spectrum of complex tumor heterogeneity and contributes to late diagnosis and treatment failure. Therefore, it is crucial to thoroughly understand the molecular mechanisms of how tumor heterogeneity develops. This review aims to summarize the possible molecular dimensions of tumor heterogeneity with an emphasis on intratumoral heterogeneity, evaluate its profound impact on the diagnosis and therapeutic strategies for HCC, and explore the suitability of appropriate pre-clinical models that can be used to best study tumor heterogeneity; thus, opening new avenues for cancer treatment.

A T cell-based SARS-CoV-2 spike protein vaccine provides protection without antibodies.

SARS-CoV-2 spike-based vaccines are used to control the COVID-19 pandemic. However, emerging variants have become resistant to antibody neutralization and further mutations may lead to full resistance. We tested whether T cells alone could provide protection without antibodies. We designed a T cell-based vaccine in which SARS-CoV-2 spike sequences were rearranged and attached to ubiquitin. Immunization of mice with the vaccine induced no specific antibodies, but strong specific T cell responses. We challenged mice with SARS-CoV-2 wild-type strain or an Omicron variant after the immunization and monitored survival or viral titers in the lungs. The mice were significantly protected against death and weight loss caused by the SARS-CoV-2 wild-type strain, and the viral titers in the lungs of mice challenged with the SARS-CoV-2 wild-type strain or the Omicron variant were significantly reduced. Importantly, depletion of CD4+ or CD8+ T cells led to significant loss of the protection. Our analyses of spike protein sequences of the variants indicated that fewer than one-third presented by dominant HLA alleles were mutated and that most of the mutated epitopes were in the subunit 1 region. As the subunit 2 region is conservative, the vaccines targeting spike protein are expected to protect against future variants due to the T cell responses.

Multiplexed discrimination of SARS-CoV-2 variants via duplex-specific nuclease combined MALDI-TOF MS.

The frequent mutations in SARS-CoV-2 significantly increase the virus's pathogenicity and transmissibility while also diminishing the effectiveness of vaccines. Consequently, assays capable of rapidly and simultaneously identifying multiple SARS-CoV-2 variants are essential for large-scale applications that aim to monitor the evolution of the virus. In this work, we propose a method combining duplex-specific nuclease (DSN)-assisted cyclic amplification with matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) detection, enabling the simultaneous identification of multiple SARS-CoV-2 variants at high-throughput. Due to the high specificity of DSN, single-base mutations can be resolved by the method. With ultra-sensitive detection by MALDI-TOF MS, a limit of detection of 100 pM viral RNA fragment was demonstrated. The assay was used for simultaneous identification and typing of SARS-CoV-2 Alpha, Beta, and Delta variants. The whole assay can be accomplished within 3 h, and the amplification is performed under constant temperature, making the technique simple in operation and efficient. It is also feasible to extend the technique to the detection of many other variants of the virus. We expect that the method can add value to the rapid screening of viral variants and can play an important role in pandemic control.

The role of magnetoencephalography in preoperative localization and postoperative outcome prediction in patients with posterior cortical epilepsy.

OBJECTIVE: We aimed to explore the value of magnetoencephalography in the presurgical evaluation of patients with posterior cortex epilepsy. METHODS: A total of 39 patients with posterior cortex epilepsy (PCE) and intact magnetoencephalography (MEG) images were reviewed from August 2019 to July 2022. MEG dipole clusters were classified into single clusters, multiple clusters, and scatter dipoles based on tightness criteria. The association of the surgical outcome with MEG dipole classifications was evaluated using Fisher's exact tests. RESULTS: Among the 39 cases, there were 24 cases of single clusters (61.5%), nine cases of multiple clusters (23.1%), and six cases of scattered dipoles (15.4%). Patients with single dipole clusters were more likely to become seizure-free. Among single dipole cluster cases (n = 24), complete MEG dipole resection yielded a more favorable surgical outcome than incomplete resection (83.3% vs. 16.7%, p = 0.007). Patients with concordant MRI and MEG findings achieved a significantly more favorable surgical outcome than discordant patients (66.7% vs. 33.3%, p = 0.044), especially in single dipole cluster patients (87.5% vs. 25.0%, p = 0.005). SIGNIFICANCE: MEG can provide additional valuable information regarding surgical candidate selection, epileptogenic zone localization, electrode implantation schedule, and final surgical planning in patients with posterior cortex epilepsy.

Weakly Supervised Lesion Detection and Diagnosis for Breast Cancers with Partially Annotated Ultrasound Images.

Deep learning (DL) has proven highly effective for ultrasound-based computer-aided diagnosis (CAD) of breast cancers. In an automatic CAD system, lesion detection is critical for the following diagnosis. However, existing DL-based methods generally require voluminous manually-annotated region of interest (ROI) labels and class labels to train both the lesion detection and diagnosis models. In clinical practice, the ROI labels, i.e. ground truths, may not always be optimal for the classification task due to individual experience of sonologists, resulting in the issue of coarse annotation to limit the diagnosis performance of a CAD model. To address this issue, a novel Two-Stage Detection and Diagnosis Network (TSDDNet) is proposed based on weakly supervised learning to improve diagnostic accuracy of the ultrasound-based CAD for breast cancers. In particular, all the initial ROI-level labels are considered as coarse annotations before model training. In the first training stage, a candidate selection mechanism is then designed to refine manual ROIs in the fully annotated images and generate accurate pseudo-ROIs for the partially annotated images under the guidance of class labels. The training set is updated with more accurate ROI labels for the second training stage. A fusion network is developed to integrate detection network and classification network into a unified end-to-end framework as the final CAD model in the second training stage. A self-distillation strategy is designed on this model for joint optimization to further improves its diagnosis performance. The proposed TSDDNet is evaluated on three B-mode ultrasound datasets, and the experimental results indicate that it achieves the best performance on both lesion detection and diagnosis tasks, suggesting promising application potential.

A critical appraisal of clinical practice guidelines on surgical treatments for spinal metastasis.

PURPOSE: As an important treatment for spinal metastasis, surgery has strict applicable conditions. Although various organizations have formulated different guidelines on surgical treatment for spinal metastasis (SM), there are certain differences in the content, standardization and quality of the guidelines and it is necessary to make a critical appraisal of them. We aim to systematically review and appraise the current guidelines on surgical treatments of SM and summarize the related recommendations with the quality evaluation of supporting evidence, as to provide a reference for the standardization of surgical treatment plans, and help clinical front-line medical workers can make safe and effective clinical decisions faster. METHODS: We searched Pubmed, Web of Science, and Embase for three major databases and online guideline databases. According to certain inclusion and exclusion criteria, the latest guidelines on the surgical treatment of SM were sorted out. AGREE II was used to evaluated the guideline's quality, and we extracted and compared the recommended treatment content of each guideline with evaluating by the evidence-grading scale. RESULTS: Eight guidelines from 2013 to 2019 were included. Seven guidelines are comprehensive guidelines and one related to the reconstructive surgery of SM. Five guidelines were evaluated as "recommended," and three guidelines were evaluated as "recommended with modifications." Regarding the indications of surgery with SM, four guidelines, seven guidelines, seven guidelines, three guidelines and three guidelines recommended surgical treatment for patients with SM with intractable pain, mechanical instability, metastatic epidural spinal cord compression (MESCC), recurrent spinal metastasis (RSM), and survival predication, respectively. Regarding the surgical strategies, three guidelines recommended minimally invasive therapy but had strict indications. Six guidelines and five guidelines recommend palliative surgery and with receiving radiation therapy, respectively. For the aggressive surgery, only one guideline recommended to apply to patients in good general conditions who has isolated symptomatic SM. Regarding the surgical reconstructions, one guideline didn't recommend iliac bone graft and three guidelines recommended PMMA bone cement. CONCLUSION: Most of the guidelines do not provide clear criteria for surgical application and provide more of a basic framework. The level of evidence for these surgical recommendations ranges from LOE B to D, and almost all guidelines recommend vertebroplasty and kyphoplasty, but for palliative and more aggressive surgery, which recommended to personalize specific surgical strategies with multidisciplinary collaboration.