Salvianolic acid B, a new type I IRE1 kinase inhibitor, abrogates AngII-induced angiogenesis by interacting with IRE1 in its active conformation.

Angiotensin II (AngII)-mediated pathological angiogenesis is one of the important factors promoting the progression of atherosclerosis, tumour metastasis, and diabetic retinopathy. Here, we first demonstrate that salvianolic acid B (Sal B) attenuated AngII-induced angiogenesis by downregulating the IRE1/ASK1/JNK/p38MAPK signalling pathway and protected vascular endothelial cells from hypoxia-induced damage. These pharmacological consequences could be ascribed to the unique interactions between Sal B and the ATP-binding cavity of IREIα, leading to bi-directional roles of IRE1 kinase and endonuclease activity; this may possibly be one of the essential mechanisms of the bi-directional regulation of angiogenesis in different conditions. Moreover, our results indicated that IRE1 was a novel anti-angiogenesis target and type I IRE1 kinase inhibitor (e.g., Sal B, APY29) and might be a potentially eligible low-toxicity drug for treating AngII-mediated pathological angiogenesis.

Longitudinal brain atlases of early developing cynomolgus macaques from birth to 48 months of age.

Longitudinal brain imaging atlases with densely sampled time-points and ancillary anatomical information are of fundamental importance in studying early developmental characteristics of human and non-human primate brains during infancy, which feature extremely dynamic imaging appearance, brain shape and size. However, for non-human primates, which are highly valuable animal models for understanding human brains, the existing brain atlases are mainly developed based on adults or adolescents, denoting a notable lack of temporally densely-sampled atlases covering the dynamic early brain development. To fill this critical gap, in this paper, we construct a comprehensive set of longitudinal brain atlases and associated tissue probability maps (gray matter, white matter, and cerebrospinal fluid) with totally 12 time-points from birth to 4 years of age (i.e., 1, 2, 3, 4, 5, 6, 9, 12, 18, 24, 36, and 48 months of age) based on 175 longitudinal structural MRI scans from 39 typically-developing cynomolgus macaques, by leveraging state-of-the-art computational techniques tailored for early developing brains. Furthermore, to facilitate region-based analysis using our atlases, we also provide two popular hierarchy parcellations, i.e., cortical hierarchy maps (6 levels) and subcortical hierarchy maps (6 levels), on our longitudinal macaque brain atlases. These early developing atlases, which have the densest time-points during infancy (to the best of our knowledge), will greatly facilitate the studies of macaque brain development.

White Matter Structural and Network Topological Changes Underlying the Behavioral Phenotype of MECP2 Mutant Monkeys.

To explore the brain structural basis underlying the behavioral abnormalities associated with Rett syndrome (RTT), we carried out detailed longitudinal noninvasive magnetic resonance imaging analyses of RTT monkey models created by gene-editing, from weaning, through adolescence, till sexual maturation. Here, we report abnormal developmental dynamics of brain white matter (WM) microstructures and network topological organizations via diffusion tensor imaging. Specifically, disrupted WM microstructural integrity was observed at 9 months, but recovered thereafter, whereas WM network topological properties showed persistent abnormal dynamics from 9 to 37 months. Changes in the WM microstructure and WM network topology were correlated well with RTT-associated behavioral abnormalities including sleep latency, environmental exploration, and conflict encounters. Deleterious and protracted early WM myelination process likely lead to abnormal synaptic pruning, resulting in poor functional segregations. Together, this study provides initial evidence for changes in WM microstructure and network topological organization, which may underlie the neuro-patho-etilogy of RTT.

Characterization of type-2 diacylglycerol acyltransferases in Haematococcus lacustris reveals their functions and engineering potential in triacylglycerol biosynthesis.

BACKGROUND: Haematococcus lacustris is an ideal source of astaxanthin (AST), which is stored in oil bodies containing esterified AST (EAST) and triacylglycerol (TAG). Diacylglycerol acyltransferases (DGATs) catalyze the last step of acyl-CoA-dependent TAG biosynthesis and are also considered as crucial enzymes involved in EAST biosynthesis in H. lacustris. Previous studies have identified four putative DGAT2-encoding genes in H. lacustris, and only HpDGAT2D allowed the recovery of TAG biosynthesis, but the engineering potential of HpDGAT2s in TAG biosynthesis remains ambiguous. RESULTS: Five putative DGAT2 genes (HpDGAT2A, HpDGAT2B, HpDGAT2C, HpDGAT2D, and HpDGAT2E) were identified in H. lacustris. Transcription analysis showed that the expression levels of the HpDGAT2A, HpDGAT2D, and HpDGAT2E genes markedly increased under high light and nitrogen deficient conditions with distinct patterns, which led to significant TAG and EAST accumulation. Functional complementation demonstrated that HpDGAT2A, HpDGAT2B, HpDGAT2D, and HpDGAT2E had the capacity to restore TAG synthesis in a TAG-deficient yeast strain (H1246) showing a large difference in enzymatic activity. Fatty acid (FA) profile assays revealed that HpDGAT2A, HpDGAT2D, and HpDGAT2E, but not HpDGAT2B, preferred monounsaturated fatty acyl-CoAs (MUFAs) for TAG synthesis in yeast cells, and showed a preference for polyunsaturated fatty acyl-CoAs (PUFAs) based on their feeding strategy. The heterologous expression of HpDGAT2D in Arabidopsis thaliana and Chlamydomonas reinhardtii significantly increased the TAG content and obviously promoted the MUFAs and PUFAs contents. CONCLUSIONS: Our study represents systematic work on the characterization of HpDGAT2s by integrating expression patterns, AST/TAG accumulation, functional complementation, and heterologous expression in yeast, plants, and algae. These results (1) update the gene models of HpDGAT2s, (2) prove the TAG biosynthesis capacity of HpDGAT2s, (3) show the strong preference for MUFAs and PUFAs, and (4) offer target genes to modulate TAG biosynthesis by using genetic engineering methods.

Cloning, expression, and characterization of a novel plant type cryptochrome gene from the green alga Haematococcus pluvialis.

A full-length cDNA sequence of plant type CRY (designated Hae-P-CRY) was cloned from the green alga Haematococcus pluvialis. The cDNA sequence was 3608 base pairs (bp) in length, which contained a 2988-bp open reading frame encoding 995 amino acids with molecular mass of 107.7 kDa and isoelectric point of 6.19. Multiple alignment analysis revealed that the deduced amino acid sequence of Hae-P-CRY shared high identity of 47-66% with corresponding plant type CRYs from other eukaryotes. The catalytic motifs of plant type CRYs were detected in the amino acid sequence of Hae-P-CRY including the typical PHR and CTE domains. Phylogenetic analysis showed that the Hae-P-CRY was grouped together with other plant type CRYs from green algae and higher plants, which distinguished from other distinct groups. The transcriptional level of Hae-P-CRY was strongly decreased after 0-4 h under HL stress. In addition, the Hae-P-CRY gene was heterologously expressed in Escherichia coli BL21 (DE3) and successfully purified. The typical spectroscopic characteristics of plant type CRYs were present in Hae-P-CRY indicated that it may be an active enzyme, which provided valuable clue for further functional investigation in the green alga H. pluvialis. These results lay the foundation for further function and interaction protein identification involved in CRYs mediated signal pathway under HL stress in H. pluvialis.

Resveratrol inhibits VEGF-induced angiogenesis in human endothelial cells associated with suppression of aerobic glycolysis via modulation of PKM2 nuclear translocation.

Endothelial cells (ECs) mainly depend on aerobic glycolysis to generate angiogenesis. Deregulation of glycolysis is often observed in human endothelial cells during angiogenesis. In the present study, we first report that resveratrol (RST), which has been intensively studied in glucose metabolism of various cancer cells, has a profound inhibitory effect on tube formation and migration via suppression of glycolysis in human umbilical vein endothelial cells (HUVECs) induced by vascular endothelial growth factor (VEGF). Moreover, we further reveal that RST reduced the mRNA and protein level of glucose transporter-1(GLUT1), hexokinase II (HK2), phosphofructokinase-1(PFK1) and pyruvate kinase M2 (PKM2) through modulation of ERK-mediated PKM2 nuclear translocation. Our results provide a novel mechanism to account for the inhibition of RST on VEGF-mediated angiogenesis and suggest that targeting aerobic glycolysis or nuclear PKM2 may be a new approach for pathological angiogenesis prevention or treatment.

BCSLinker: automatic method for constructing a knowledge graph of venous thromboembolism based on joint learning.

Background: Venous thromboembolism (VTE) is characterized by high morbidity, mortality, and complex treatment. A VTE knowledge graph (VTEKG) can effectively integrate VTE-related medical knowledge and offer an intuitive description and analysis of the relations between medical entities. However, current methods for constructing knowledge graphs typically suffer from error propagation and redundant information. Methods: In this study, we propose a deep learning-based joint extraction model, Biaffine Common-Sequence Self-Attention Linker (BCSLinker), for Chinese electronic medical records to address the issues mentioned above, which often occur when constructing a VTEKG. First, the Biaffine Common-Sequence Self-Attention (BCsSa) module is employed to create global matrices and extract entities and relations simultaneously, mitigating error propagation. Second, the multi-label cross-entropy loss is utilized to diminish the impact of redundant information and enhance information extraction. Results: We used the electronic medical record data of VTE patients from a tertiary hospital, achieving an F1 score of 86.9% on BCSLinker. It outperforms the other joint entity and relation extraction models discussed in this study. In addition, we developed a question-answering system based on the VTEKG as a structured data source. Conclusion: This study has constructed a more accurate and comprehensive VTEKG that can provide reference for diagnosing, evaluating, and treating VTE as well as supporting patient self-care, which is of considerable clinical value.

Different MRI-based radiomics models for differentiating misdiagnosed or ambiguous pleomorphic adenoma and Warthin tumor of the parotid gland: a multicenter study.

Purpose: To evaluate the effectiveness of MRI-based radiomics models in distinguishing between Warthin tumors (WT) and misdiagnosed or ambiguous pleomorphic adenoma (PA). Methods: Data of patients with PA and WT from two centers were collected. MR images were used to extract radiomic features. The optimal radiomics model was found by running nine machine learning algorithms after feature reduction and selection. To create a clinical model, univariate logistic regression (LR) analysis and multivariate LR were used. The independent clinical predictors and radiomics were combined to create a nomogram. Two integrated models were constructed by the ensemble and stacking algorithms respectively based on the clinical model and the optimal radiomics model. The models' performance was evaluated using the area under the curve (AUC). Results: There were 149 patients included in all. Gender, age, and smoking of patients were independent clinical predictors. With the greatest average AUC (0.896) and accuracy (0.839) in validation groups, the LR model was the optimal radiomics model. In the average validation group, the radiomics model based on LR did not have a higher AUC (0.795) than the clinical model (AUC = 0.909). The nomogram (AUC = 0.953) outperformed the radiomics model in terms of discrimination performance. The nomogram in the average validation group had a highest AUC than the stacking model (0.914) or ensemble model (0.798). Conclusion: Misdiagnosed or ambiguous PA and WT can be non-invasively distinguished using MRI-based radiomics models. The nomogram exhibited excellent and stable diagnostic performance. In daily work, it is necessary to combine with clinical parameters for distinguishing between PA and WT.

Diagnostic value of circulating miR-155 for breast cancer: a meta-analysis.

Backgrounds: The value of circulating microRNA (miR)-155 for breast cancer (BC) diagnosis may differ in different studies. Therefore, we conducted this systematic review and meta-analysis to evaluate the potential application of circulating miR-155 in the diagnosis of BC. Methods: Articles published before December 2023 and in English were searched in these databases: PubMed, Web of Science, Medline, EMBASE and Google Scholar. A summary of sensitivity, specificity, positive likelihood ratios (PLR), negative likelihood ratios (NLR), and diagnostic odds ratio (DOR) were calculated from the true positive (TP), true negative (TN), false positive (FP) and false negative (FN) of each study. Additionally, the summary receive-operating characteristics (SROC) curve was constructed to summarize the TP and FP rates. Results: The pooled parameters calculated were as follows: sensitivity, 0.93 (95% CI: 0.83-0.97); specificity, 0.85 (95% CI: 0.74-0.92); PLR, 6.4 (95% CI: 3.4-11.9); NLR, 0.09 (95% CI: 0.04-0.20); and DOR, 74 (95% CI: 22-247). The analysis showed a significant heterogeneity (sensitivity, I2 = 95.19%, p < 0.001; specificity, I2 = 95.29%, p < 0.001; DOR, I2 = 92.9%, p < 0.001). The SROC curve was with an area under curve (AUC) of 0.95 (95% CI: 0.93-0.97). Conclusion: Circulating miR-155 has a potential in the diagnosis of BC.

Spatial distribution and environmental correlations of Culex pipiens pallens (Diptera: Culicidae) in Haidian district, Beijing.

Culex pipiens pallens Coquillett, 1898 (Diptera: Culicidae) was the dominant health threat to mosquito species in Beijing, and it is important to unravel the spatial distribution and environmental correlations of Cx. pipiens pallens in Beijing. 3S technology methods and spatial statistics were used to clarify the distribution pattern. Subsequently, linear and spatial regression were performed to detect the environmental factors linked with the density of Cx. pipiens pallens. The same "middle peak" spatial distribution pattern was observed for Cx. pipiens pallens density at the community, subdistrict, and loop area levels in our study area. In addition, there were various correlated environmental factors at the community and subdistrict scales. At the community scale, the summary values of the Modified Normalized Difference Water Index (MNDWI) in 2 km buffer zone (MNDWI_2K) were negatively correlated, and the summary values of Normalized Difference Built-up Index (NDBI) in 800 m buffer zone (NDBI_800) was positively correlated to the Cx. pipiens pallens density. However, the summary values of Normalized Difference Vegetation Index and Nighttime Light Index significantly affected Cx. pipiens pallens density at the subdistrict scale. Our findings provide insight into the spatial distribution pattern of Cx. pipiens pallens density and its associated environmental risk factors at different spatial scales in the Haidian district of Beijing for the first time. The results could be used to predict the Cx. pipiens pallens density as well as the risk of lymphatic filariasis (LF) infection, which would help implement prevention and control measures to prevent future risks of biting and LF transmission in Beijing.

Key function of Kouleothrix in stable formation of filamentous aerobic granular sludge at low superficial gas velocity with polymeric substrates.

Forming and maintaining stable aerobic granular sludge (AGS) at a low superficial gas velocity (SGV) is challenging, particularly with polymeric substrates. This study cultivated filamentous aerobic granular sludge (FAGS) with filamentous Kouleothrix (Type 1851) at low SGV (0.15 cm/s) utilizing mixed acetate-soluble starch. Within approximately 260 days, notable increases in the relative abundance of Kouleothrix (from 4 % to 10 %) and Ca. Competibacter (from 1 % to 26 %) were observed through 16S rRNA gene analysis. Metagenomic analysis revealed increased expression of functional genes involved in volatile fatty acid (VFA) production (e.g., ackA and pta) and polyhydroxyalkanoate synthesis (e.g., phbB and phbC). Kouleothrix acted as a skeleton for bacterial attachment and was the key fermenting bacteria promoting granulation and maintaining granule stability. This study provides insight into the formation of FAGS with low-energy and non-VFA substrates.

Automating venous thromboembolism risk assessment: a dual-branch deep learning method using electronic medical records.

Background: Venous thromboembolism (VTE) is a prevalent cardiovascular disease. Although risk assessment and preventive measures are effective, manual assessment is inefficient and covers a small population in clinical practice. Hence, it is necessary to explore intelligent methods for VTE risk assessment. Methods: The Padua scale has been widely used in VTE risk assessment, and we divided its assessment into disease category judgment and comprehensive clinical information judgment according to the characteristics of the Padua scale. We proposed a dual-branch deep learning (DB-DL) assessment method. First, in the disease category branch, we propose a deep learning-based Padua disease classification model (PDCM) for determining patients' Padua disease categories by considering patients' diagnosis, symptoms, and symptom weights. In the branch of comprehensive clinical information, we use the Chinese lexical analysis (LAC) word separation technique, combined with professional corpus and rules, to extract and judge the comprehensive clinical factors in the electronic medical record (EMR). Results: We validated the accuracy of the method with the Padua assessment results of 7,690 Chinese clinical EMRs. First, our proposed method allows for a fully automated assessment, and the average time to assess one patient is only 0.37 s. Compared to the gold standard, our method has an Area Under Curve (AUC) value of 0.883, a specificity value of 0.957, and a sensitivity value of 0.816 for assessing the Padua risk patient class. Conclusion: Our DB-DL assessment method automates VTE risk assessment, thereby addressing the challenges of time-consuming evaluation and limited population coverage. Thus, this method is highly clinically valuable.

Optimized Radiomics Nomogram Based on Automated Breast Ultrasound System: A Potential Tool for Preoperative Prediction of Metastatic Lymph Node Burden in Breast Cancer.

Background: Axillary lymph node dissection (ALND) can be safely avoided in women with T1 or T2 primary invasive breast cancer (BC) and one to two metastatic sentinel lymph nodes (SLNs). However, cancellation of ALND based solely on SLN biopsy (SLNB) may lead to adverse outcomes. Therefore, preoperative assessment of LN tumor burden becomes a new focus for ALN status. Objective: This study aimed to develop and validate a nomogram incorporating the radiomics score (rad-score) based on automated breast ultrasound system (ABUS) and other clinicopathological features for evaluating the ALN status in patients with early-stage BC preoperatively. Methods: Totally 354 and 163 patients constituted the training and validation cohorts. They were divided into ALN low burden (<3 metastatic LNs) and high burden (≥3 metastatic LNs) based on the histopathological diagnosis. The radiomics features of the segmented breast tumor in ABUS images were extracted and selected to generate the rad-score of each patient. These rad-scores, along with the ALN burden predictors identified from the clinicopathologic characteristics, were included in the multivariate analysis to establish a nomogram. It was further evaluated in the training and validation cohorts. Results: High ALN burdens accounted for 11.2% and 10.8% in the training and validation cohorts. The rad-score for each patient was developed based on 7 radiomics features extracted from the ABUS images. The radiomics nomogram was built with the rad-score, tumor size, US-reported LN status, and ABUS retraction phenomenon. It achieved better predictive efficacy than the nomogram without the rad-score and exhibited favorable discrimination, calibration and clinical utility in both cohorts. Conclusion: We developed an ABUS-based radiomics nomogram for the preoperative prediction of ALN burden in BC patients. It would be utilized for the identification of patients with low ALN burden if further validated, which contributed to appropriate axillary treatment and might avoid unnecessary ALND.

Bioimprinted lipase-catalyzed synthesis of medium- and long-chain structured lipids rich in docosahexaenoic acid for infant formula.

Medium- and long-chain structured lipids (MLSLs) rich in docosahexaenoic acid (DHA) were obtained in shorter reaction time by acidolysis of single-cell oil (DHASCO) from Schizochytrium sp. with caprylic acid (CA) using a lipase bioimprinted with fatty acids as a catalyst. The conditions for preparation of the bioimprinted lipase for the acidolysis reaction were firstly optimized and the activity of the obtained lipase was 2.17 times higher than that of the non-bioimprinted. The bioimprinted lipase was then used as a catalyst and the reaction conditions were optimized. Under the optimal conditions, the equilibrium could be achieved in 4 h, and the total and sn-1,3 CA contents in the product were 29.18% and 42.34%, respectively, and the total and sn-2 DHA contents were 46.26% and 70.12%, respectively. Such MLSLs rich in sn-1,3 CA and sn-2 DHA are beneficial for DHA absorption, and thus have potential for use in infant formula.

Implementation of renormalized excitonic method at ab initio level.

The renormalized excitonic method [Hajj et al., Phys Rev B 2005, 72, 224412], in which the excited state of the whole system may be described as a linear combination of local excitations, has been implemented at ab initio level. Its performance is tested on the ionization potential and the energy gap between singlet ground state and lowest triplet for linear molecular hydrogen chains and more realistic systems, such as polyenes and polysilenes, using full configuration interaction (FCI) wave functions with a minimal basis set. The influence of different block sizes and the extent of interblock interactions are investigated. It has been demonstrated that satisfactory results can be obtained if the near degeneracies between the model space and the outer space are avoided and if interactions between the next-nearest neighbor blocks are considered. The method can be used with larger basis sets and other accurate enough ab initio evaluations (instead of FCI) of local excited states, from blocks, or from dimers or trimers of blocks. It provides a new possibility to accurately and economically describe the low-lying delocalized excited states of large systems, even inhomogeneous ones.

Predictive value of T2-weighted imaging and dynamic contrast-enhanced MRI for assessing cervical invasion in patients with endometrial cancer: a meta-analysis.

PURPOSE: To obtain the diagnostic accuracy of T2-weighted imaging (T2WI), and dynamic contrast-enhanced MRI (DCE-MRI) in the preoperative assessment of cervical invasion in patients with endometrial cancer (EC). METHODS: Databases including PubMed, Embase, Cochrane Library, Web of Science, and Clinical Trials were searched for relevant articles published from January 2000 to August 2020. Pooled estimation data were obtained by statistical analysis. RESULTS: In total, 24 articles were included. For assessing cervical invasion of EC, the pooled sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic odds ratio (DOR), and area under the curve (AUC) for T2WI were 0.70 (0.61-0.77), 0.92 (0.89-0.94), 8.7 (6.5-11.6), 0.33 (0.25-0.43), 26 (17-41), and 0.92 (0.89-0.94), respectively. For DCE-MRI, the pooled sensitivity, specificity, PLR, NLR, DOR, and AUC were 0.75 (0.60-0.85), 0.95 (0.89-0.98), 14.7 (6.6-32.9), 0.27 (0.16-0.44), 55 (18-165), and 0.92 (0.89-0.94), respectively; for T2WI combined with DCE-MRI, they were 0.58 (0.41-0.73), 0.98 (0.95-0.99), 28.1 (12.8-62.1), 0.43 (0.30-0.63), 65 (29-146), and 0.94 (0.91-0.96), respectively. CONCLUSIONS: DCE-MRI demonstrated higher diagnostic performance than T2WI in the prediction of cervical invasion in patients with EC. T2WI combined with DCE-MRI improved the pooled specificity, PLR, DOR, and AUC compared to T2WI alone or DCE-MRI alone.

Improved version of a local contracted configuration interaction of singles and doubles with partial inclusion of triples and quadruples.

A local contracted single and double configuration interaction (LC-CISD) method, which introduces contracted singly and doubly excited vectors within the framework of bond functions, has been recently proposed [P. Reinhardt et al., J. Chem. Phys. 129, 164106 (2008)]. The present work improves this method by introducing a coupled-electron pair approximation (CEPA-3) dressing and by incorporating the leading part of linked effects of triples (T) and quadruples (Q) through a series of local four-electron full CI calculations. Two different ways have been adopted to incorporate this linked TQ effect. One consists of dressing the first column/line of the whole LC-CISD matrix. The other one introduces an additional contracted wave function responsible for the linked effect for each bond pair. The present LC-CEPA-3+TQ treatments have been applied to the evaluation of equilibrium bond lengths and harmonic frequencies of diatomic molecules (HF, BF, CuH, N(2), F(2), and Cl(2)) and single bond breaking in HF, CH(4), ClCH(3), ClSiH(3), n-butane, and F(2) molecules, symmetrical stretching of the two OH bonds in a water molecule, and symmetrical expansion of a triangular Be(3) cluster. The results show that the performance of the LC-CEPA-3+TQs compares favorably with coupled-cluster singles and doubles (CCSD) and CCSD(T) methods, presenting similar behaviors around equilibrium and better ones for stretched geometries. The LC-CEPA-3 method is strictly separable, and the size consistency error of our treatment of triples and quadruples is extremely small. The strict separability can be further achieved by dressing the doubly excited bond functions with the linked TQ effect. The efficiency of truncations on the bielectronic integrals has also been tested.

Nomogram Including Elastography for Prediction of Contralateral Central Lymph Node Metastasis in Solitary Papillary Thyroid Carcinoma Preoperatively.

BACKGROUND: It is controversial whether contralateral prophylactic central neck dissection (PCND) should be performed for patients with solitary and clinical lymph node negative (cN0) papillary thyroid carcinoma (PTC) although routine ipsilateral PCND is required. OBJECTIVE: The aim of this study was to develop an improved nomogram including clinical features, ultrasound, and acoustic radiation force impulse (ARFI) elastography for the prediction of contralateral central lymph node metastasis (CLNM) in patients with solitary and cN0 PTC in the preoperative period. MATERIALS AND METHODS: A total of 340 patients were retrospectively included as the training cohort and 170 patients as the external validation cohort. Patients were grouped according to the pathological results of contralateral CLNM. The association between the clinical characteristics, ultrasound, and ARFI elastography and the risk for contralateral CLNM were analyzed. A nomogram was established based on the result of multivariable logistic analysis to predict the risk of contralateral CLNM, which was assessed by internal and external validation. RESULTS: CLNM was found in 213 patients (41.8%), among whom 142 (27.8%) had ipsilateral CLNM and 95 (18.6%) had contralateral CLNM (including 68 (13.3%) with bilateral CLNM). Multivariable analysis revealed that patients with younger age, male gender, larger tumor size, closer distance from the capsule, microcalcification, and larger SWVmean were independent predictors associated with the contralateral CLNM (P < 0.05), which was served as the basis of the nomogram. It showed good discrimination (C-index: 0.856) and calibration (χ2 = 9.028, P = 0.340, Hosmer-Lemeshow test) in the training cohort, and good discrimination was maintained in the external validation cohort (C-index: 0.792). CONCLUSION: The nomogram utilizing the features of ultrasound combined with ARFI elastography in preoperatively predicting the risk of contralateral CLNM in patients with solitary and cN0 PTC was established, which showed superior performance both in internal and external validation.

Unsupervised learning of probabilistic object models (POMs) for object classification, segmentation, and recognition using knowledge propagation.

We present a method to learn probabilistic object models (POMs) with minimal supervision, which exploit different visual cues and perform tasks such as classification, segmentation, and recognition. We formulate this as a structure induction and learning task and our strategy is to learn and combine elementary POMs that make use of complementary image cues. We describe a novel structure induction procedure, which uses knowledge propagation to enable POMs to provide information to other POMs and "teach them" (which greatly reduces the amount of supervision required for training and speeds up the inference). In particular, we learn a POM-IP defined on Interest Points using weak supervision [1], [2] and use this to train a POM-mask, defined on regional features, which yields a combined POM that performs segmentation/localization. This combined model can be used to train POM-edgelets, defined on edgelets, which gives a full POM with improved performance on classification. We give detailed experimental analysis on large data sets for classification and segmentation with comparison to other methods. Inference takes five seconds while learning takes approximately four hours. In addition, we show that the full POM is invariant to scale and rotation of the object (for learning and inference) and can learn hybrid objects classes (i.e., when there are several objects and the identity of the object in each image is unknown). Finally, we show that POMs can be used to match between different objects of the same category, and hence, enable objects recognition.

Two-dimensional multilabel active learning with an efficient online adaptation model for image classification.

Conventional active learning dynamically constructs the training set only along the sample dimension. While this is the right strategy in binary classification, it is suboptimal for multilabel image classification. We argue that for each selected sample, only some effective labels need to be annotated while others can be inferred by exploring the label correlations. The reason is that the contributions of different labels to minimizing the classification error are different due to the inherent label correlations. To this end, we propose to select sample-label pairs, rather than only samples, to minimize a multilabel Bayesian classification error bound. We call it two-dimensional active learning because it considers both the sample dimension and the label dimension. Furthermore, as the number of training samples increases rapidly over time due to active learning, it becomes intractable for the offline learner to retrain a new model on the whole training set. So we develop an efficient online learner to adapt the existing model with the new one by minimizing their model distance under a set of multilabel constraints. The effectiveness and efficiency of the proposed method are evaluated on two benchmark data sets and a realistic image collection from a real-world image sharing Web site-Corbis.