Design and implementation of railway green performance basic data collection system

This study addressed the issues related to the collection and management of basic data for railway green performance. A railway green performance basic database has been constructed based on metadata and data exchange schemas. A data classification system has been established from the perspectives of businesses, processes,and entities. A BIM(Building Information Modelling) model data extraction scheme is proposed based on field similarity matching and a document content extraction scheme is proposed based on image recognition. A railway green performance basic data collection system has been developed, achieving efficient collection and integrated management of railway green performance basic data. This system can provide data support for applications such as railway carbon emissions accounting, green cost-benefit analysis, and evaluation of green design solutions.

Classifying coke using CT scans and landmark multidimensional scaling

One factor that limits development of fundamental research on the infuence of coke microstructure on its strength is the difculty in quantifying the way that microstructure is both classifed and distributed in three dimensions.To support such fundamental studies,this study evaluated a novel volumetric approach for classifying small(approx.450μm^(3))blocks of coke microstructure from 3D computed tomography scans.An automated process for classifying microstructure blocks was described.It is based on Landmark Multi-Dimensional Scaling and uses the Bhattacharyya metric and k-means clustering.The approach was evaluated using 27 coke samples across a range of coke with diferent properties and reliably identifed 6 ordered class of coke microstructure based on the distribution of voxel intensities associated with structural density.The lower class(1–2)subblocks tend to be dominated by pores and thin walls.Typically,there is an increase in wall thickness and reduced pore sizes in the higher classes.Inert features are also likely to be seen in higher classes(5–6).In general,this approach provides an efcient automated means for identifying the 3D spatial distribution of microstructure in CT scans of coke.

Gallbladder perforation:A single-center experience in north India and a step-up approach for management

Background: Spontaneous gallbladder perforation(GBP) is an uncommon diagnosis. This study presented the experience of managing spontaneous GBP over nine years at a large, tertiary care university hospital in north India and investigated the outcomes and treatment strategies. Methods: A retrospective review of prospectively maintained digital database of consecutive patients was performed. All patients received medical and/or surgical treatment for spontaneous GBP in our department between January 2010 and June 2018. Results: We identifed 151 patients(81 females and 70 males) with mean age of 53 years. Most common presenting features were pain(96.7%), fever(54.3%) and jaundice(31.1%). Most common cause was gallbladder stones(84.8%) followed by common bile duct stones(30.5%), xanthogranulomatous cholecystitis(17.9%) and malignancy(11.9%). As per Niemeier classifcation, 8.6% had type 1 GBP(free perforation in peritoneal cavity), 76.2% had type 2 GBP(localized perforation) and 13.2% had type 3 GBP(cholecystoenteric fstula). About 60% of the perforations were diagnosed preoperatively. Type 1 was more common in patients with diabetes and also had the worst prognosis. Surgery was performed in 109 patients(72.2%). Seven patients(4.6%) had a postoperative morbidity of Clavien-Dindo Ⅲ or higher. There were three mortalities in patients who underwent surgery. Conclusions: High index of suspicion is required for preoperative diagnosis of GBP, especially in types 2 and 3. Laparoscopic cholecystectomy can be difficult in these patients and patients may require open or partial cholecystectomy. Early diagnosis and step-up approach for the treatment of GBP is critical.

Systematic Review on Wearable Lower Extremity Robotic Exoskeletons for Assisted Locomotion

Lower extremity robotic exoskeletons(LEEX)can not only improve the ability of the human body but also provide healing treatment for people with lower extremity dysfunction.There are a wide range of application needs and development prospects in the military,industry,medical treatment,consumption and other felds,which has aroused widespread concern in society.This paper attempts to review LEEX technical development.First,the history of LEEX is briefy traced.Second,based on existing research,LEEX is classifed according to auxiliary body parts,structural forms,functions and felds,and typical LEEX prototypes and products are introduced.Then,the latest key technologies are analyzed and summarized,and the research contents,such as bionic structure and driving characteristics,human–robot interaction(HRI)and intentawareness,intelligent control strategy,and evaluation method of power-assisted walking efciency,are described in detail.Finally,existing LEEX problems and challenges are analyzed,a future development trend is proposed,and a multidisciplinary development direction of the key technology is provided.

Improving the classifcation accuracy using biomarkers selected from machine learning methods

High-dimensional data encountered in genomic and proteomic studies are often limited by the sample size but has a higher number of predictor variables.Therefore selecting the most relevant variables that are correlated with the outcome variable is a crucial step.This paper describes an approach for selecting a set of optimal variables to achieve a classifcation model with high predictive accuracy.The work described using a biological classifer published elsewhere but it can be generalized for any application.

Genomic diversity and ecological distribution of marine Pseudoalteromonas phages

Pseudoalteromonas,with a ubiquitous distribution,is one of the most abundant marine bacterial genera.It is especially abundant in the deep sea and polar seas,where it has been found to have a broad metabolic capacity and unique co-existence strategies with other organisms.However,only a few Pseudoalteromonas phages have so far been isolated and investigated and their genomic diversity and distribution patterns are still unclear.Here,the genomes,taxonomic features and distribution patterns of Pseudoalteromonas phages are systematically analyzed,based on the microbial and viral genomes and metagenome datasets.A total of 143 complete or nearly complete Pseudoalteromonas-associated phage genomes(PSAPGs)were identifed,including 34 Pseudoalteromonas phage isolates,24 proviruses,and 85 Pseudoalteromonas-associated uncultured viral genomes(UViGs);these were assigned to 47 viral clusters at the genus level.Many integrated proviruses(n=24)and flamentous phages were detected(n=32),suggesting the prevalence of viral lysogenic life cycle in Pseudoalteromonas.PSAPGs encoded 66 types of 249 potential auxiliary metabolic genes(AMGs)relating to peptidases and nucleotide metabolism.They may also participate in marine biogeochemical cycles through the manipulation of the metabolism of their hosts,especially in the phosphorus and sulfur cycles.Siphoviral and flamentous PSAPGs were the predominant viral lineages found in polar areas,while some myoviral and siphoviral PSAPGs encoding transposase were more abundant in the deep sea.This study has expanded our understanding of the taxonomy,phylogenetic and ecological scope of marine Pseudoalteromonas phages and deepens our knowledge of viral impacts on Pseudoalteromonas.It will provide a baseline for the study of interactions between phages and Pseudoalteromonas in the ocean.

Transfer active learning by querying committee

In real applications of inductive learning for classifcation,labeled instances are often defcient,and labeling them by an oracle is often expensive and time-consuming.Active learning on a single task aims to select only informative unlabeled instances for querying to improve the classifcation accuracy while decreasing the querying cost.However,an inevitable problem in active learning is that the informative measures for selecting queries are commonly based on the initial hypotheses sampled from only a few labeled instances.In such a circumstance,the initial hypotheses are not reliable and may deviate from the true distribution underlying the target task.Consequently,the informative measures will possibly select irrelevant instances.A promising way to compensate this problem is to borrow useful knowledge from other sources with abundant labeled information,which is called transfer learning.However,a signifcant challenge in transfer learning is how to measure the similarity between the source and the target tasks.One needs to be aware of diferent distributions or label assignments from unrelated source tasks;otherwise,they will lead to degenerated performance while transferring.Also,how to design an efective strategy to avoid selecting irrelevant samples to query is still an open question.To tackle these issues,we propose a hybrid algorithm for active learning with the help of transfer learning by adopting a divergence measure to alleviate the negative transfer caused by distribution diferences.To avoid querying irrelevant instances,we also present an adaptive strategy which could eliminate unnecessary instances in the input space and models in the model space.Extensive experiments on both the synthetic and the real data sets show that the proposed algorithm is able to query fewer instances with a higher accuracy and that it converges faster than the state-of-the-art methods.

Roadmap toward subtype-specifc vulnerabilities in adult glioma

Gliomas are the most common primary malignancies in the adult central nervous system(CNS),and over the course of the last decades a wealth of data on their genomic characterization has been acquired.Nevertheless,attempts to stratify patients on the basis of this work has so far conspicuously failed to identify useful treatment targets,and no phase III clinical trials conducted to date have reached a favorable outcome.We suggest that these translational failures are due to inadequacies in classifcation schemes,which fail to capture the range of biologically distinct entities that give rise to gliomas.Treating gliomas of diferent subtypes together,rather than as a set of biologically distinct but related tumors,has resulted in a classifcation scheme rich in unexplained heterogeneities,and has restricted target identifcation eforts to cell cycle and cell growth regulators.We suggest that this failure of detailed genomic characterizations to identify useful treatment targets requires a re-assessment of our assumptions concerning glioma origins.We propose a re-interpretation of glioma subtypes in the light of knowledge of the developmental pathways of the various neural lineages that make up the adult CNS.Such a developmental subtype-specifc classifcation scheme based on dys-regulated cell fate decisions may not only improve classifcation and diagnosis but,more importantly,identify potentially druggable subtype-specifc developmental vulnerabilities.