Search

Computational prediction of ligand-protein binding is an essential part of overcoming the synthesis bottleneck for rapid development of small-molecule drugs. Nevertheless, an accurate prediction of binding free energy is challenging without first establishing a ligand binding pose. Reported in this paper is an advanced optimization algorithm that is capable of providing multiple, high-fidelity solutions to a ligand-protein pose even for complicated systems with large numbers of degrees of freedom. This algorithm achieves high performance by incorporating several important features, such as niching and force-field annealing. Results from several challenging use cases are presented and discussed.

3D geographical information systems (GIS) software tools that support synchronous collaboration efforts among distributed decision-making participants can be very useful in many application areas, such as urban planning, engineering design, disaster and emergency management, and subsurface resources management. Although research has been extensively carried out in related fields such as groupware system, collaborative virtual environments and collaborative industrial design, a comprehensive study has not been found in the 3D GIS field. The scientific goal of this research is to add synchronous collaboration capability to the Internet-enabled 3D GIS environment. More specifically, the ojbective of this research is to investigate methods and key technologies to design a synchronous collaborative Internet-enabled 3D GIS environment (SC3DGIS). In this thesis, two basic questions for synchronous collaborative 3D GIS are presented: 1) What are the special functions and usability for SC3DGIS compared to traditional and mainstream GIS applications? 2) What are the special problems in design and development of SC3DGIS when considering it as a specific type of synchronous collaborative system? In answering the above two questions, a conceptual framework is developed to investigate main aspects which play core roles in reflecting the features of SC3DGIS. The more detailed system requirements are further analyzed through a case study. The prototype design adopts two layer structures - the shared 3D environment layer and the agent assistant layer. The shared 3D environment adopts a semi-replicated architecture, while in the agent assistant layer, a multi-agent method is used to solve the complex interactions between users and the shared understanding of all parties in the framework. Two walkthroughs are presented to validate the usability of the prototype. The results of this research indicate that 1) adding synchronous and collaborative capabilities to a 3D GIS environment can significantly improve the efficiency and satisfaction of decision-making for the geographically-distributed people; 2) the events transferring method and semi-replicated architecture are more suitable for synchronous collaborative 3D GIS than a display imagery transferring method and centralized architecture because of the highly interactive 3D contents; 3) a 3D data model for collaborative purposes require explicit data state presentations, such as color and style; 4) a social collaboration model and its ontology presentation and shared catching method keep the distributed system consistent and understandable; and 5) a multi-agent method sitting on a shared 3D view can be very helpful in assisting group users to carry out more complex communications, such as discussions and making deals.

Geographical information systems (GIS) software tools that support synchronous collaboration efforts among distributed decision-making participants can be very useful in many application areas, such as urban planning, engineering design, disaster and emergency response, and distant learning. However, most existing GIS tools do not provide adequate support for group interaction on decision-making and design scenarios. Early efforts on developing collaborative GIS tools have focused on collaborative geospatial information sharing and presentation in a group environment, mostly adapted to centralized client-server architecture for specific applications. This thesis presents the results of a research project, aiming at providing such GIS software tools over the Internet. Based on the analysis of two mainstream architectures used in collaborative applications: centralized architecture and replicated architecture, a hybrid architecture is selected to develop a collaborative GIS framework as the platform for prototyping the aforementioned GIS tools. The discussion focuses on synchronous collaboration where people interact with each other using the system at the same time from different places. The prototype system, called GeoLink, addresses some important design and development issues such as session management and floor control through a message sending approach.

OBJECTIVE Stereotactic radiosurgery (SRS) has been increasingly employed in recent years to treat intracranial metastatic lesions. However, there is still a need for optimization of treatment paradigms to provide better local control and prevent progressive intracranial disease. In the current study, the authors utilized a national collaborative registry to investigate the outcomes of patients with intracranial metastatic disease who underwent SRS and to determine factors associated with lesion treatment response, overall progression, and mortality. METHODS The NeuroPoint Alliance SRS registry was queried for all patients with intracranial metastatic lesions undergoing single- or multifraction SRS at participating institutions between 2016 and 2020. The main outcomes of interest included lesion response (lesion-level analysis), progression using Response Assessment for Neuro-Oncology criteria, and mortality (patient-level analysis). Kaplan-Meier analysis was used to report time to progression and overall survival, and multivariable Cox proportional hazards analysis was used to investigate factors associated with lesion response, progression, and mortality. RESULTS A total of 501 patients (1447 intracranial metastatic lesions) who underwent SRS and had available follow-up were included in the current analyses. The most common primary tumor was lung cancer (49.5%, n = 248), followed by breast (15.4%, n = 77) and melanoma (12.2%, n = 61). Most patients had a single lesion (44.9%, n = 225), 29.3% (n = 147) had 2 or 3 lesions, and 25.7% (n = 129) had > 3 lesions. The mean sum of baseline measurements of the lesions according to Response Evaluation Criteria in Solid Tumors (RECIST) was 35.54 mm (SD 25.94). At follow-up, 671 lesions (46.4%) had a complete response, 631 (43.6%) had a partial response (≥ 30% decrease in longest diameter) or were stable (< 30% decrease but < 20% increase), and 145 (10%) showed progression (> 20% increase in longest diameter). On multivariable Cox proportional hazards analysis, melanoma-associated lesions (HR 0.48, 95% CI 0.34–0.67; p < 0.001) and larger lesion size (HR 0.94, 95% CI 0.93–0.96; p < 0.001) showed lower odds of lesion regression, while a higher biologically effective dose was associated with higher odds (HR 1.001, 95% CI 1.0001–1.00023; p < 0.001). A total of 237 patients (47.3%) had overall progression (local failure or intracranial progressive disease), with a median time to progression of 10.03 months after the index SRS. Factors found to be associated with increased hazards of progression included male sex (HR 1.48, 95% CI 1.108–1.99; p = 0.008), while administration of immunotherapy (before or after SRS) was found to be associated with lower hazards of overall progression (HR 0.62, 95% CI 0.460–0.85; p = 0.003). A total of 121 patients (23.95%) died during the follow-up period, with a median survival of 19.4 months from the time of initial SRS. A higher recursive partitioning analysis score (HR 21.3485, 95% CI 1.53202–3.6285; p < 0.001) was found to be associated with higher hazards of mortality, while single-fraction treatment compared with hypofractionated treatment (HR 0.082, 95% CI 0.011–0.61; p = 0.015), administration of immunotherapy (HR 0.385, 95% CI 0.233–0.64; p < 0.001), and presence of single compared with > 3 lesions (HR 0.427, 95% CI 0.187–0.98; p = 0.044) were found to be associated with lower risk of mortality. CONCLUSIONS The comparability of results between this study and those of previously published clinical trials affirms the value of multicenter databases with real-world data collected without predetermined research purpose.

This retrospective cohort study aims to determine the proportion of postmenopausal women with recurrent urinary tract infections (rUTIs) who improve with vaginal estrogen cream alone and to identify risk factors for those who go on to require additional therapies.Postmenopausal women presenting with rUTIs seen at a single institution over a 29-month period were identified. Patients who had not undergone previous treatment and were started on vaginal estrogen cream alone as first-line therapy were included in the study. Descriptive statistics were used to compare demographic and clinical characteristics between therapy groups. Log-binomial regression was used to explore the association between treatment failure (ie, the need for additional therapy for rUTIs) and patient sociodemographic and clinical characteristics.During the study period, 167 patients with rUTIs met inclusion criteria. Of these, 67.7% noted improvement or resolution in their symptoms with vaginal estrogen cream alone. Women with a concomitant diagnosis of urinary incontinence, as defined as subjective report of bothersome stress or urgency urinary incontinence, were 2.3 times more likely to need additional therapy compared with women not reporting urinary incontinence (relative risk, 2.28; 95% confidence interval, 1.06-4.90).Sixty-eight percent of postmenopausal women with rUTIs treated with vaginal estrogen alone did not go on to receive additional therapy. A risk factor for needing additional therapy in our study population is the presence of urinary incontinence.

Radiology report generation aims at generating descriptive text from radiology images automatically, which may present an opportunity to improve radiology reporting and interpretation. A typical setting consists of training encoder-decoder models on image-report pairs with a cross entropy loss, which struggles to generate informative sentences for clinical diagnoses since normal findings dominate the datasets. To tackle this challenge and encourage more clinically-accurate text outputs, we propose a novel weakly supervised contrastive loss for medical report generation. Experimental results demonstrate that our method benefits from contrasting target reports with incorrect but semantically-close ones. It outperforms previous work on both clinical correctness and text generation metrics for two public benchmarks., Findings of EMNLP 2021

SUMMARYAnimals with complex nervous systems demand sleep for memory consolidation and synaptic remodeling. Here we show that though theCaenorhabditis elegansnervous system has a limited number of neurons, sleep is necessary for both processes. In addition, it is unclear in any system if sleep collaborates with experience to alter synapses between specific neurons and whether this ultimately affects behavior.C. elegansneurons have defined connections and well-described contributions to behavior. We show that spaced odor-training and post-training sleep induce long-term memory. Memory consolidation, but not acquisition, requires a pair of interneurons, the AIYs, which play a role in odor-seeking behavior. In worms that consolidate memory, both sleep and odor conditioning are required to diminish inhibitory synaptic connections between the AWC chemosensory neurons and the AIYs. Thus, we demonstrate in a living organism that sleep is required for events immediately after training that drive memory consolidation and alter synaptic structures.

Background: Di(2-ethylhexyl)phthalate (DEHP) is one of the phthalates most widely used to manufacture various plastic products. However, the potential effects of DEHP on erythropoiesis have not been investigated comprehensively. Here, we aimed to investigate whether DEHP modulate the function of haematopoietic stem and progenitor cells (HSPCs) to influence erythropoiesis, and to explore the associated mechanisms.Results: In the present study, human cell lines with a capacity to differentiate into erythroid cells and murine bone marrow cells were treated with DEHP. DEHP not only impaired HSPC function but also suppressed erythroid differentiation in a dose-dependent manner. In addition, DEHP removal restored HSPC activity. To explore how DEHP interferes with erythroid differentiation, we focused on energy metabolism and Klotho expression. DEHP suppressed erythroid differentiation via upregulation of Klotho expression in erythroblasts and not via cellular bioenergetic modulation.Conclusions: Here, we have demonstrated a novel pathophysiological role of DEHP in erythroid differentiation. DEHP impaired erythroid differentiation by inducing Klotho expression and not by reprogramming cellular bioenergetic profiles during differentiation.

As of 28 August 2020, there have been 5.88 million Coronavirus Disease 2019 (COVID19) cases and 181,000 COVID-19 related deaths in the United States alone. Given the lack of an effective pharmaceutical treatment for COVID-19, the high contagiousness of the disease and its varied clinical outcomes, identifying patients at risk of progressing to severe disease is crucial for the allocation of valuable healthcare resources during this pandemic. Current research has shown that there is a higher prevalence of cardiovascular comorbidities amongst patients with severe COVID-19 or COVID-19-related deaths, but the link between cardiovascular disease and poorer prognosis is poorly understood. We believe that pre-existing immune dysregulation that accompanies cardiovascular disease predisposes patients to a harmful inflammatory immune response, leading to their higher risk of severe disease. Thus, in this project, we aim to characterize immune dysregulation in patients with cardiomyopathy, venous thromboembolism and COVID-19 patients by looking at immune-associated gene dysregulation, immune infiltration and dysregulated immunological pathways and gene signatures.

Purpose:While the double crush phenomena (compression along two points on a nerve) has been established between median neuropathy and cervical radiculopathy, combined suprascapular neuropathy (SSN) and cervical C5/C6 radiculopathy—so-called shoulder double crush syndrome—has not been well examined. We aim to identify the incidence of shoulder double crush syndrome in patients undergoing arthroscopic suprascapular nerve release for SSN.Methods:One hundred consecutive patients >18 years of age who were positive for SSN on electromyography and motor nerve conduction studies (EMG/NCS) and underwent a suprascapular nerve release were included. Patients with evidence of shoulder double crush syndrome were identified based on x ray, cervical spine magnetic resonance imaging (MRI) and examination findings. Demographics, electrodiagnostics results, treatment courses, and clinical outcomes (visual analog scores and rotator cuff strength) following arthroscopic suprascapular nerve release were compared between patients with double crush syndrome versus isolated SSN.Results:Thirty one percent of patients had evidence of shoulder crush syndrome. Two significant electrophysiologic differences were noted in shoulder double crush patients compared to isolated SSN patients. Patients with double crush had an increased incidence of median neuropathy (51% vs 30%, P = .04). Double crush patients had less supraspinatus motor amplitude difference between the affected side and non-affected side compared to isolated SSN patients (2.62 mV vs 3.44 mV, P = .03). In general, most double crush patients were treated conservatively with regard to their cervical spine pathology.Conclusion:A significant percentage of patients with SSN have evidence of shoulder double crush syndrome. Patients with SSN and concomitant median neuropathy should have a detailed neck examination performed.

The website and database https://snengs.nichd.nih.gov provides RNA sequencing data from multi-species analysis of the pineal glands from zebrafish (Danio rerio), chicken (White Leghorn), rat (Rattus novegicus), mouse (Mus musculus), rhesus macaque (Macaca mulatta), and human (Homo sapiens); in most cases, retinal data are also included along with results of the analysis of a mixture of RNA from tissues. Studies cover day and night conditions; in addition, a time series over multiple hours, a developmental time series and pharmacological experiments on rats are included. The data have been uniformly re-processed using the latest methods and assemblies to allow for comparisons between experiments and to reduce processing differences. The website presents search functionality, graphical representations, Excel tables, and track hubs of all data for detailed visualization in the UCSC Genome Browser. As more data are collected from investigators and improved genomes become available in the future, the website will be updated. This database is in the public domain and elements can be reproduced by citing the URL and this report. This effort makes the results of 21st century transcriptome profiling widely available in a user-friendly format that is expected to broadly influence pineal research. Fil: Chang, Eric. National Instituto of Child Health & Human Development; Estados Unidos Fil: Fu, Cong. National Instituto of Child Health & Human Development; Estados Unidos Fil: Coon, Steven L.. National Instituto of Child Health & Human Development; Estados Unidos Fil: Alon, Shahar. No especifíca; Fil: Bozinoski, Marjan. No especifíca; Fil: Breymaier, Matthew. National Instituto of Child Health & Human Development; Estados Unidos Fil: Bustos, Diego Martin. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - Mendoza. Instituto de Histología y Embriología de Mendoza Dr. Mario H. Burgos. Universidad Nacional de Cuyo. Facultad de Ciencias Médicas. Instituto de Histología y Embriología de Mendoza Dr. Mario H. Burgos; Argentina Fil: Clokie, Samuel J.. National Instituto of Child Health & Human Development; Estados Unidos Fil: Gothilf, Yoav. No especifíca; Fil: Esnault, Caroline. National Instituto of Child Health & Human Development; Estados Unidos Fil: Iuvone, P. Michael. Emory University School of Medicine; Estados Unidos Fil: Mason, Christopher E.. No especifíca; Fil: Ochocinska, Margaret J.. National Instituto of Child Health & Human Development; Estados Unidos Fil: Tovin, Adi. No especifíca; Fil: Wang, Charles. Loma Linda University; Estados Unidos Fil: Xu, Pinxian. No especifíca; Fil: Zhu, Jinhang. No especifíca; Fil: Dale, Ryan. National Instituto of Child Health & Human Development; Estados Unidos Fil: Klein, David C.. National Instituto of Child Health & Human Development; Estados Unidos

Fine-grained object recognition and classification in biomedical images poses a number of challenges. Images typically contain multiple instances (e.g. glands) and the recognition of salient structures is confounded by visually complex backgrounds. Due to the cost of data acquisition or the limited availability of specimens, data sets tend to be small. We propose a simple yet effective attention based deep architecture to address these issues, specially to improve background suppression and recognition of important instances per image. Attention maps per instance are learnt in an end-to-end fashion. Microscopic images of fungi (new data) and a publicly available Breast Cancer Histology benchmark dataset are used to demonstrate the performance of the proposed approach. Experimental results suggest that the proposed approach advances the state-of-the-art.

Background Vancomycin area under the concentration-time curve (AUC) has been linked to efficacy and safety. An accurate method of calculating the AUC is needed. Methods Bayesian dose-optimizing software programs available for clinician use and first-order pharmacokinetic equations were evaluated for their ability to estimate vancomycin AUC. A previously published rich pharmacokinetic data set of 19 critically ill patients was used for validation of the AUC estimation. The AUC estimated using subsets of the full data set by Bayesian software and equations was compared with the reference AUC. Accuracy (ratio of estimated AUC to the reference AUC) and bias (percentage difference of estimated AUC to reference AUC) were calculated. Results Five Bayesian dose-optimizing software programs (Adult and Pediatric Kinetics [APK], BestDose, DoseMe, InsightRx, and Precise PK) and two first-order pharmacokinetic equations were included. Of the Bayesian programs, InsightRx was the most adaptable, visually appealing, easiest to use, and had the most company support. Utilizing only the trough, accuracy (range 0.79-1.03) and bias (range 5.1-21.2%) of the Bayesian dose-optimizing software were variable. Precise PK and BestDose had the most accurate estimates with the accuracy values of BestDose exhibiting the most variability of all the programs; however, both programs were more difficult to use. Precise PK was the least biased (median 5.1%). Using a single nontrough value produced similar results to that of the trough for most programs. The addition of a second level to the trough improved the accuracy and bias for DoseMe and InsightRx but not Precise PK and BestDose. APK did not reliably estimate the AUC with input of two levels. Using two levels, the pharmacokinetic equations produced similar or better accuracy and bias as compared with Bayesian software. Conclusion Bayesian dose-optimizing software using one or more vancomycin levels and pharmacokinetic equations utilizing two vancomycin levels produce similar estimates of the AUC.

Introduction: Venous thromboembolism and in-situ small vessel thrombosis are increased in hospitalized patients with COVID-19 in several patient cohorts. Endotheliopathy and activation of both platelets and coagulation predict critical illness and death. For these reasons the use of anti-platelet agents and increased-intensity anticoagulation in the care of hospitalized patients with COVID-19 is under intense study in several clinical trials. We sought to examine the impact of aspirin and anticoagulation on hospitalization outcomes. Methods: We examined outcomes in a large multi-site cohort of consecutive, hospitalized, COVID-19 laboratory confirmed patients under a risk-stratified treatment algorithm from March 13 through June 18, with a focus on efficacy of aspirin and/or increased-intensity anticoagulation. Out of 4150 identified hospitalized patients with COVID-19, we created 3 study cohorts. The overall cohort (2785 patients) excluded pediatric patients, those with incomplete electronic data, and those with multiple admissions. The aspirin (1956 patients) and anticoagulation (1623 patients) cohorts were nested within the overall cohort; the former excluded patients on any home anti-platelet therapy or those who received non-aspirin anti-platelet therapy in the hospital, while the latter excluded patients who did not receive prophylactic or intermediate dose anticoagulation in the hospital. The primary outcome was in-hospital death. Secondary outcomes were time-to-death with a competing risk (time-to-hospital-discharge), escalation to ICU, length-of-stay and use of mechanical ventilation. Variables examined included age, gender, BMI, race, Rothman Index (RI), D-dimer (DD) and patient co-morbidities including cardiovascular disease, chronic kidney disease, and prior VTE. The aspirin and anticoagulation cohorts underwent propensity score (PS) matching utilizing variables found to be significant in multivariable regression modeling in the overall cohort with 638 and 386 patients, respectively. Results: Univariate followed by multivariable regression modeling in the 2785 patient overall cohort established a novel role for RI, and independent roles for age, BMI, and maximum DD, in predicting severity of illness. In all cohorts the 50th and lower percentile of admission RI was predictive of mortality in multivariable modeling (i.e. aspirin: 3rd and 4th admission RI quartiles with HR = 0.18 for both, p Conclusion: We show in a large cohort of consecutively hospitalized patients with COVID-19 treated under a risk-stratified algorithm the prognostic utility of the admission RI in assessing outcomes in hospitalized patients with COVID-19 and a potential benefit of aspirin therapy on in-hospital death from COVID-19. A potential albeit marginal benefit of intermediate dose anticoagulation over prophylactic dose anticoagulation merits further study with results of clinical trials awaited. Figure Disclosures Neuberg: Pharmacyclics: Research Funding; Madrigak Pharmaceuticals: Current equity holder in publicly-traded company; Celgene: Research Funding.

In this paper, we propose an innovative end-to-end subtitle detection and recognition system for videos in East Asian languages. Our end-to-end system consists of multiple stages. Subtitles are firstly detected by a novel image operator based on the sequence information of consecutive video frames. Then, an ensemble of Convolutional Neural Networks (CNNs) trained on synthetic data is adopted for detecting and recognizing East Asian characters. Finally, a dynamic programming approach leveraging language models is applied to constitute results of the entire body of text lines. The proposed system achieves average end-to-end accuracies of 98.2% and 98.3% on 40 videos in Simplified Chinese and 40 videos in Traditional Chinese respectively, which is a significant outperformance of other existing methods. The near-perfect accuracy of our system dramatically narrows the gap between human cognitive ability and state-of-the-art algorithms used for such a task.

Ceftaroline is often used in durations greater than that studied in clinical trials. Several retrospective, non-comparative studies have suggested a higher than anticipated incidence of neutropenia in patients receiving prolonged treatment with ceftaroline. We sought to determine if ceftaroline was associated with neutropenia by comparing the incidence with ceftaroline treatment with treatment with several comparative antibiotics.Patients receiving 14 or more consecutive days of treatment with ceftaroline were compared with patients receiving cefazolin, daptomycin, linezolid, nafcillin or vancomycin (control group). The primary outcome was the development of neutropenia. Multivariate logistic regression and propensity score weighting using inverse probability weights with regression adjustment were used to control for confounding variables.A total of 753 patients were included (53 that received ceftaroline and 700 that received a comparative antibiotic). Ceftaroline was associated with a greater incidence of neutropenia as compared with the control group (17.0% versus 3.9%, P 0.001). Several covariates were also associated with neutropenia and included younger age, lower baseline absolute neutrophil count, liver disease and bone and joint infections. After controlling for these confounders, receipt of ceftaroline continued to be associated with the development of neutropenia (adjusted OR 3.97, P = 0.003). Analysis after propensity score weighting confirmed this finding.The results of this study suggest that prolonged treatment with ceftaroline is associated with a greater incidence of neutropenia as compared with other antibiotics that are often used for treatment of staphylococcal infections. Careful monitoring of absolute neutrophil count is recommended in patients receiving14 days of ceftaroline.

This paper proposes a multi-level feature learning framework for human action recognition using a single body-worn inertial sensor. The framework consists of three phases, respectively designed to analyze signal-based (low-level), components (mid-level) and semantic (high-level) information. Low-level features capture the time and frequency domain property while mid-level representations learn the composition of the action. The Max-margin Latent Pattern Learning (MLPL) method is proposed to learn high-level semantic descriptions of latent action patterns as the output of our framework. The proposed method achieves the state-of-the-art performances, 88.7%, 98.8% and 72.6% (weighted F 1 score) respectively, on Skoda, WISDM and OPP datasets.

Despite limited clinical data, ceftaroline is commonly used for treatment of complicated, invasive infections caused by methicillin-resistant Staphylococcus aureus (MRSA). A retrospective chart review was conducted of adult patients receiving ceftaroline for MRSA osteomyelitis admitted between April 2011 and March 2016 at a five-hospital system. Twelve patients met the inclusion criteria. All patients received prior antimicrobial therapy with a median time to switch to ceftaroline of 45.5 days. Five of the 12 patients (41.7%) met criteria for ceftaroline failure. Patients with vertebral osteomyelitis (58%) had a longer length of stay, longer ceftaroline treatment, but similar success rates to those with non-vertebral osteomyelitis (57% vs. 60%). Ceftaroline is a viable alternative for a challenging patient population that has failed or are unable to receive other therapies.

DNA methylation, catalyzed by DNA methyltransferases (DNMTs), is a heritable epigenetic mark, participating in numerous physiological processes. DNMT3A is of particular relevance to hematopoietic differentiation, because DNMT3A mutations are strongly related to hematopoietic malignancies. Additionally, DNMT3A deficiency has been reported to increase the hematopoietic stem cell pool by limiting their differentiation. Our previous study demonstrated that complete loss of DNMT3A resulted in anemia, while DNMT3A haploinsufficiency caused an elevated population of erythrocytes in the content of oncogenic KRAS. Since erythropoiesis is tightly regulated via the erythropoietin (EPO)-mediated RAS-RAF-MEK-ERK1/2 pathway, the question arises whether DNMT3A cooperates with RAS signaling to modulate erythropoiesis. Human leukemia cell lines were used, with differentiation capabilities towards megakaryocyte and erythroid lineages. Overexpression of DNMT3A was found to enhance erythrocytic differentiation of K562 cells, while DNMT3A knockdown suppressed differentiation. Furthermore, higher DNMT3A expression was detected in late-stage mouse erythroblasts along with the DNMT3A translocation to the nucleus. Further studies demonstrated that both ERK1/2-DNMT3A interaction and serine-255 phosphorylation in DNMT3A led to DNMT3A translocation into the nucleus, and modulated erythrocytic differentiation. Our results not only explore the critical role of DNMT3A in erythropoiesis, but also provide a new insight into ERK1/2-DNMT3A interaction in the hematopoietic system.

Fine-grained classification of microscopic image data with limited samples is an open problem in computer vision and biomedical imaging. Deep learning based vision systems mostly deal with high number of low-resolution images, whereas subtle details in biomedical images require higher resolution. To bridge this gap, we propose a simple yet effective deep network that performs two tasks simultaneously in an end-to-end manner. First, it utilises a gated attention module that can focus on multiple key instances at high resolution without extra annotations or region proposals. Second, the global structural features and local instance features are fused for final image level classification. The result is a robust but lightweight end-to-end trainable deep network that yields state-of-the-art results in two separate fine-grained multi-instance biomedical image classification tasks: a benchmark breast cancer histology dataset and our new fungi species mycology dataset. In addition, we demonstrate the interpretability of the proposed model by visualising the concordance of the learned features with clinically relevant features.

Automatic Non-rigid Histological Image Registration (ANHIR) challenge was organized to compare the performance of image registration algorithms on several kinds of microscopy histology images in a fair and independent manner. We have assembled 8 datasets, containing 355 images with 18 different stains, resulting in 481 image pairs to be registered. Registration accuracy was evaluated using manually placed landmarks. In total, 256 teams registered for the challenge, 10 submitted the results, and 6 participated in the workshop. Here, we present the results of 7 well-performing methods from the challenge together with 6 well-known existing methods. The best methods used coarse but robust initial alignment, followed by non-rigid registration, used multiresolution, and were carefully tuned for the data at hand. They outperformed off-the-shelf methods, mostly by being more robust. The best methods could successfully register over 98% of all landmarks and their mean landmark registration accuracy (TRE) was 0.44% of the image diagonal. The challenge remains open to submissions and all images are available for download.

Background Consisting of dictated free-text documents such as discharge summaries, medical narratives are widely used in medical natural language processing. Relationships between anatomical entities and human body parts are crucial for building medical text mining applications. To achieve this, we establish a mapping system consisting of a Wikipedia-based scoring algorithm and a named entity normalization method (NEN). The mapping system makes full use of information available on Wikipedia, which is a comprehensive Internet medical knowledge base. We also built a new ontology, Tree of Human Body Parts (THBP), from core anatomical parts by referring to anatomical experts and Unified Medical Language Systems (UMLS) to make the mapping system efficacious for clinical treatments.Result The gold standard is derived from 50 discharge summaries from our previous work, in which 2,224 anatomical entities are included. The F1-measure of the baseline system is 70.20%, while our algorithm based on Wikipedia achieves 86.67% with the assistance of NEN.Conclusions We construct a framework to map anatomical entities to THBP ontology using normalization and a scoring algorithm based on Wikipedia. The proposed framework is proven to be much more effective and efficient than the main baseline system.

Background Quantitative areas is of great measurement of wound significance in clinical trials, wound pathological analysis, and daily patient care. 2D methods cannot solve the problems caused by human body curvatures and different camera shooting angles. Our objective is to simply collect wound areas, accurately measure wound areas and overcome the shortcomings of 2D methods. Results We propose a method with 3D transformation to measure wound area on a human body surface, which combines structure from motion (SFM), least squares conformal mapping (LSCM), and image segmentation. The method captures 2D images of wound, which is surrounded by adhesive tape scale next to it, by smartphone and implements 3D reconstruction from the images based on SFM. Then it uses LSCM to unwrap the UV map of the 3D model. In the end, it utilizes image segmentation by interactive method for wound extraction and measurement. Our system yields state-of-the-art results on a dataset of 118 wounds on 54 patients, and performs with an accuracy of 0.97. The Pearson correlation, standardized regression coefficient and adjusted R square of our method are 0.999, 0.895 and 0.998 respectively. Conclusions A smartphone is used to capture wound images, which lowers costs, lessens dependence on hardware, and avoids the risk of infection. The quantitative calculation of the 3D wound area is realized, solving the challenges that 2D methods cannot and achieving a good accuracy.

Category: Ankle, Ankle Arthritis, Hindfoot, Midfoot/Forefoot Introduction/Purpose: Augment is the first Food and Drug Administration approved synthetic bone substitute specifically formulated for foot and ankle procedures and have shown results comparable to autograft. Initial studies on the use of Augment for fusion procedures of the foot and ankle are encouraging but there remains a lack of radiographic outcomes on its use. The purposes of this study were to report on the radiographic fusion rate of Augment and identify risk factors for nonunion in patients treated with Augment. Methods: After institutional review board approval, a retrospective study was performed of all patients ages 18-85 who underwent arthrodesis of the ankle, hindfoot, or midfoot and used Augment alone or in combination with autograft and/or allograft. Patients were excluded if they did not have regular radiographic follow-up. Each operative report was examined for procedure performed and location of the foot involved. Post-operative radiographs were reviewed to look at union rate following surgery. Patient charts were reviewed for age, gender, Body Mass Index (BMI), smoking history, and diabetes to examine potential risk factors for nonunion. Continuous variables were reported in means ± standard deviations and differences determined between those that fused and those that did not using a Wilcoxon Rank Sums test. Categorical variables reported as number of patients (percent of patients) and differences determined between those that fused and those that did not using a Fisher’s exact test. Results: A total of 71 patients (average age at injury 57±14 years; 35 males (49%), 36 females (50%); BMI 31±6) underwent 33 (46%) midfoot arthrodesis, 41 (58%) hindfoot arthrodesis, and 27 (38%) ankle arthrodesis. A total of 58 patients (82%) went on to achieve fusion while 13 (18%) did not. There was no difference in gender (p=0.135), age (p=0.345) or BMI (p=0.196) between those who achieved fusion and those who did not. Patients with diabetes had a greater risk of nonunion compared to those who did not (p=0.033) while current smoking status or revision fusion did not pose a risk factor (p=1.000) (Table 1). There were no differences in part of the foot/ankle involved (p=0.445) or bone graft used (p=0.303). Conclusion: This the first study to examine the radiographic follow up of union using Augment. The most important finding of this study is that the rate of fusion in foot and ankle arthrodesis using Augment is near 82%. Diabetes is a risk factor for nonunion when using Augment. Smoking, gender, age, BMI, autograft vs allograft, and part of the foot or ankle fused did not pose a risk factor.

Deep learning (DL) has achieved state-of-the-art performance in many digital pathology analysis tasks. Traditional methods usually require hand-crafted domain-specific features, and DL methods can learn representations without manually designed features. In terms of feature extraction, DL approaches are less labor intensive compared with conventional machine learning methods. In this paper, we comprehensively summarize recent DL-based image analysis studies in histopathology, including different tasks (e.g., classification, semantic segmentation, detection, and instance segmentation) and various applications (e.g., stain normalization, cell/gland/region structure analysis). DL methods can provide consistent and accurate outcomes. DL is a promising tool to assist pathologists in clinical diagnosis.

Tarsal tunnel syndrome (TTS) is a relatively uncommon compression neuropathy caused by impingement of the tibial nerve or one of the terminal branches. The presence of accessory musculature at the posteromedial aspect of the ankle has been identified as a rare cause of this condition. Despite the rarity of this condition, it must be considered in patients with refractory symptoms consistent with tibial nerve dysfunction. The accurate diagnosis of this condition relies heavily on a detailed history and physical examination, adequate imaging read by both surgeon and trained musculoskeletal radiologist, as well as a high level of suspicion for such pathology. In this case report, we describe a 46-year-old male with history, examination, and imaging all consistent with TTS secondary to accessory musculature. Following excision of an accessory soleus and flexor digitorum accessorius longus, as well as simultaneous tarsal tunnel release, the patient experienced full resolution of his symptoms. This highlights the importance of considering accessory musculature as a potential cause of TTS in patients presenting with tibial compression neuropathy. Levels of Evidence: Level V: Case Report

Purpose: In this paper, the authors proposed a new 3D registration algorithm, 3D-scale invariant feature transform (SIFT)-Flow, for multiatlas-based liver segmentation in computed tomography (CT) images. Methods: In the registration work, the authors developed a new registration method that takes advantage of dense correspondence using the informative and robust SIFT feature. The authors computed the dense SIFT features for the source image and the target image and designed an objective function to obtain the correspondence between these two images. Labeling of the source image was then mapped to the target image according to the former correspondence, resulting in accurate segmentation. In the fusion work, the 2D-based nonparametric label transfer method was extended to 3D for fusing the registered 3D atlases. Results: Compared with existing registration algorithms, 3D-SIFT-Flow has its particular advantage in matching anatomical structures (such as the liver) that observe large variation/deformation. The authors observed consistent improvement over widely adopted state-of-the-art registration methods such as ELASTIX, ANTS, and multiatlas fusion methods such as joint label fusion. Experimental results of liver segmentation on the MICCAI 2007 Grand Challenge are encouraging, e.g., Dice overlap ratio 96.27% ± 0.96% by our method compared with the previous state-of-the-art result of 94.90% ± 2.86%. Conclusions: Experimental results show that 3D-SIFT-Flow is robust for segmenting the liver from CT images, which has large tissue deformation and blurry boundary, and 3D label transfer is effective and efficient for improving the registration accuracy.

Brain dysfunction, including neuroinflammation, metabolic changes, and cognitive impairment has been reported in sepsis and other disorders “traditionally” characterized by peripheral immune and metabolic dysregulation (Annu Rev Immunol, 2018, 36:783–812). In this context, non-invasive evaluation of neuroinflammation is important for strategizing new diagnostic and therapeutic approaches targeting the brain. We utilized Micro Positron Emission Tomography (microPET) with [18F]Fluoro-2-deoxy-2-D-glucose (18FDG) and 11C-Peripheral Benzodiazepine Receptor ([11C]PBR) to evaluate brain metabolic alterations and microglia activation (indicating neuroinflammation) during murine endotoxemia. Lipopolysaccharide (LPS, endotoxin) (2 mg/kg) or saline was injected (i.p.) in male C57Bl/6 mice 6 hrs prior to microPET imaging. [11C] PBR28 (~0.5 mCi) was injected via the tail vein followed by 60 mins dynamic imaging. Subsequently, 18FDG (0.5 mCi) was injected i.p. with a 10 mins static scan acquired following 40 mins uptake. Brain images were analyzed in a standard anatomical space using Statistical Parametric Mapping. Significantly (P

3D medical image registration is of great clinical importance. However, supervised learning methods require a large amount of accurately annotated corresponding control points (or morphing), which are very difficult to obtain. Unsupervised learning methods ease the burden of manual annotation by exploiting unlabeled data without supervision. In this paper, we propose a new unsupervised learning method using convolutional neural networks under an end-to-end framework, Volume Tweening Network (VTN), for 3D medical image registration. We propose three innovative technical components: (1) An end-to-end cascading scheme that resolves large displacement; (2) An efficient integration of affine registration network; and (3) An additional invertibility loss that encourages backward consistency. Experiments demonstrate that our algorithm is 880x faster (or 3.3x faster without GPU acceleration) than traditional optimization-based methods and achieves state-of-theart performance in medical image registration., Comment: Journal of Biomedical and Health Informatics

Tumor proliferation is an important biomarker indicative of the prognosis of breast cancer patients. Assessment of tumor proliferation in a clinical setting is highly subjective and labor-intensive task. Previous efforts to automate tumor proliferation assessment by image analysis only focused on mitosis detection in predefined tumor regions. However, in a real-world scenario, automatic mitosis detection should be performed in whole-slide images (WSIs) and an automatic method should be able to produce a tumor proliferation score given a WSI as input. To address this, we organized the TUmor Proliferation Assessment Challenge 2016 (TUPAC16) on prediction of tumor proliferation scores from WSIs. The challenge dataset consisted of 500 training and 321 testing breast cancer histopathology WSIs. In order to ensure fair and independent evaluation, only the ground truth for the training dataset was provided to the challenge participants. The first task of the challenge was to predict mitotic scores, i.e., to reproduce the manual method of assessing tumor proliferation by a pathologist. The second task was to predict the gene expression based PAM50 proliferation scores from the WSI. The best performing automatic method for the first task achieved a quadratic-weighted Cohen's kappa score of $\kappa$ = 0.567, 95% CI [0.464, 0.671] between the predicted scores and the ground truth. For the second task, the predictions of the top method had a Spearman's correlation coefficient of r = 0.617, 95% CI [0.581 0.651] with the ground truth. This was the first study that investigated tumor proliferation assessment from WSIs. The achieved results are promising given the difficulty of the tasks and weakly-labelled nature of the ground truth. However, further research is needed to improve the practical utility of image analysis methods for this task., Comment: Overview paper of the TUPAC16 challenge: http://tupac.tue-image.nl/

To determine inter-lab reliability in sleep stage scoring using the 2014 American Academy of Sleep Medicine (AASM) manual. To understand in-depth reasons for disagreement and provide suggestions for improvement. This study consisted of 40 all-night polysomnographys (PSGs) from different samples. PSGs were segmented into 37,642 30-s epochs. Five doctors from China and two doctors from America scored the epochs following the 2014 AASM standard. Scoring disagreement between two centers was evaluated using Cohen’s kappa (κ). After visual inspection of PSGs of deviating scorings, potential disagreement reasons were analyzed. Inter-lab reliability yielded a substantial degree (κ = 0.75 ± 0.01). Scoring for stage W (κ = 0.89) and R (κ = 0.87) achieved the highest agreement, while stage N1 (κ = 0.45) reflected the lowest. Considering the relative disagreement ratio, N2-N3 (22.09%), W-N1 (19.68%), and N1-N2 (18.75%) were the most frequent combinations of discrepancy. American and Chinese doctors showed certain characteristics in the scoring of discrepancy combination W-N1, N1-N2, and N2-N3. There are seven reasons for disagreement, namely “on-threshold characteristic” (29.21%), “context influence” (18.06%), “characteristic identification difficulty” (8.81%), “arousal-wake confusion” (7.57%), “derivation inconsistence” (2.15%), “on-borderline characteristic” (0.92%), and “misrecognition” (33.27%). This study demonstrated the sleep stage scoring agreement of the 2014 AASM manual and explored potential sources of labeling ambiguity. Improvement measures were suggested accordingly to help remove ambiguity for scorers and improve scoring reliability at the international level.

INTRODUCTION: We present a case describing the management of a woman with severe, functionally limiting cervical myeloradiculopathy in the setting of congenital cervical canal stenosis and Arnold–Chiari I malformation. CASE PRESENTATION: The subject is a 57-year-old woman with prior anterior cervical discectomy and fusion who presented with left-sided neck pain associated with radiculopathy, migraine, gait incoordination, and cervical dystonia. Cervical stenosis and Chiari malformation were confirmed using MRI. Conservative management with botulinum toxin, oral muscle relaxants, and cervical brace led to gradual exacerbation of symptoms. Due to failure of conservative management, surgical decompression with C3–C6 posterior laminoplasty was performed, resulting in complete resolution of all symptoms and markedly improved quality of life. DISCUSSION: This case reports a severe and nonspecific presentation of cervical myeloradiculopathy. Surgery for cervical myeloradiculopathy is controversial, and conservative therapy is initially preferred. However, in this case, conservative treatments likely led to paraspinal weakness, cervical hypermobility, and biomechanical instability, resulting in exacerbation of symptoms. Stretch/shear forces have been postulated to accelerate cervical myelopathy, and excessive cervical instability and range of motion are significant predictors of deterioration. In this case, surgical decompression with posterior cervical laminoplasty after 1 year of conservative management yielded significant pain relief and functional restoration, indicating the utility of this procedure even in the presence of Arnold–Chiari I malformation. This case illustrates that decompression can be effective for refractory cervical myeloradiculopathy associated with Chiari malformation, congenital stenosis, and prior anterior instrumentation, and highlights the potential risks of prolonged conservative management.

This study compared and analyzed both the energy savings and cost-effectiveness of various energy-saving technologies retrofitted to common buildings in China. Base models for an office and store building, set in representative climate zones of China—Beijing, Shanghai, and Guangzhou—were established and calibrated in EnergyPlus, a building energy simulation software program. Various energy-saving technologies were then applied to these models in EnergyPlus to calculate the overall energy savings under different climate conditions. In addition, a payback analysis was performed to determine the cost-effectiveness of each technology. The final results of this study can serve as a preliminary reference for selecting effective and economical energy-saving technologies to retrofit existing public buildings.

In this paper, we tackle the problem of common object (multiple classes) discovery from a set of input images, where we assume the presence of one object class in each image. This problem is, loosely speaking, unsupervised since we do not know a priori about the object type, location, and scale in each image. We observe that the general task of object class discovery in a fully unsupervised manner is intrinsically ambiguous; here we adopt saliency detection to propose candidate image windows/patches to turn an unsupervised learning problem into a weakly-supervised learning problem. In the paper, we propose an algorithm for simultaneously localizing objects and discovering object classes via bottom-up (saliency-guided) multiple class learning (bMCL). Our contributions are three-fold: (1) we adopt saliency detection to convert unsupervised learning into multiple instance learning, formulated as bottom-up multiple class learning (bMCL); (2) we propose an integrated framework that simultaneously performs object localization, object class discovery, and object detector training; (3) we demonstrate that our framework yields significant improvements over existing methods for multi-class object discovery and possess evident advantages over competing methods in computer vision. In addition, although saliency detection has recently attracted much attention, its practical usage for high-level vision tasks has yet to be justified. Our method validates the usefulness of saliency detection to output "noisy input" for a top-down method to extract common patterns.

Rich pharmacokinetic data on vancomycin in critically ill patients are lacking. The purpose of this study was to evaluate the pharmacokinetics of vancomycin in this population using rich pharmacokinetic sampling. Nineteen critically ill patients received individualized vancomycin doses by intermittent infusion to achieve target trough concentrations (15–20 mg/L). Blood samples were collected following the third or later dose of vancomycin. Serial blood samples were collected at 30 min following initiation of the vancomycin infusion; at the end of the infusion; serially at 60, 120, 300, and 480 min after the infusion finished; and immediately prior to the next dose. Vancomycin concentration–time profiles at steady state were fit to a noncompartmental model to determine the pharmacokinetic parameters. Vancomycin trough concentration was correlated to AUC0–24 (r = 0.83, P Vancomycin trough concentration may not be an adequate surrogate of AUC0–24. Additional research is needed to determine dosing strategies to optimize AUC0–24 while limiting toxicity.