A hybrid knowledge and ensemble classification approach for prediction of venous thromboembolism

Clinical narratives such as progress summaries, lab reports, surgical reports, and other narrative texts contain key biomarkers about a patient's health. Evidence-based preventive medicine needs accurate semantic and sentiment analysis to extract and classify medical features as the input to appropriate machine learning classifiers. However, the traditional approach of using single classifiers is limited by the need for dimensionality reduction techniques, statistical feature correlation, a faster learning rate, and the lack of consideration of the semantic relations among features. Hence, extracting semantic and sentiment-based features from clinical text and combining multiple classifiers to create an ensemble intelligent system overcomes many limitations and provides a more robust prediction outcome. The selection of an appropriate approach and its interparameter dependency becomes key for the success of the ensemble method. This paper proposes a hybrid knowledge and ensemble learning framework for prediction of venous thromboembolism (VTE) diagnosis consisting of the following components: a VTE ontology, semantic extraction and sentiment assessment of risk factor framework, and an ensemble classifier. Therefore, a component-based analysis approach was adopted for evaluation using a data set of 250 clinical narratives where knowledge and ensemble achieved the following results with and without semantic extraction and sentiment assessment of risk factor, respectively: a precision of 81.8% and 62.9%, a recall of 81.8% and 57.6%, an F measure of 81.8% and 53.8%, and a receiving operating characteristic of 80.1% and 58.5% in identifying cases of VTE.     

Sarcoplasmic reticular Ca2+ pump ATPase activity in congestive heart failure due to myocardial infarction

OBJECTIVE: Earlier studies have shown a depression in the sarcoplasmic reticular (SR) Ca2+ uptake and gene expression in Ca2+ pump ATPase protein in congestive heart failure subsequent to myocardial infarction. It is the objective of this study to understand further the mechanisms of depressed SR Ca2+ pump activity in the failing heart. METHODS: Heart failure in rats was induced by occluding the left coronary artery for 16 weeks and the viable left ventricle was processed for the isolation of SR membranes. Sham-operated animals were used as control. The characteristics of SR Ca2+ pump ATPase in the presence of different concentrations of K+, Ca2+ and ATP were examined and the purity of these membranes was monitored by determining the marker enzyme activities. In addition to measuring changes in cyclic adenosine monophosphate (cAMP) protein kinase and Ca(2+)-calmodulin induced phosphorylation, alterations in SR phospholipid composition as well as sulfhydryl (SH) group content were investigated. RESULTS: Ca(2+)-stimulated ATPase activity, unlike Mg(2+)-ATPase activity, was depressed in the left ventricular SR from failing hearts as compared to control. The decrease in Ca(2+)-stimulated ATPase activity was seen at different concentrations of Ca2+, K+ and ATP but no changes in the affinities of the enzyme for Ca2+ and ATP were evident. The SR Ca(2+)-stimulated ATPase activities in the presence of both cAMP-dependent protein kinase and Ca(2+)-calmodulin were markedly decreased in the failing hearts when compared to control preparations. Furthermore, the 32P incorporation in the presence of cAMP-dependent protein kinase or Ca(2+)-calmodulin was also reduced in the experimental heart SR membranes. The phospholipid composition of the SR membranes from the failing heart was markedly altered. No changes in SH-group or the degree of cross contamination with other membranes were apparent in the failing heart SR. CONCLUSIONS: These results suggest that abnormalities in membrane phospholipid composition and phosphorylation of the enzyme may partly explain the observed depression in SR Ca2+ pump ATPase activity in heart failure following myocardial infarction.     

Autonomous mapping of HL7 RIM and relational database schema

Healthcare systems need to share information within and across the boundaries in order to provide better care to the patients. For this purpose, they take advantage of the full potential of current state of the art in healthcare standards providing interoperable solutions. HL7 V3 specification is an international message exchange and interoperability standard. HL7 V3 messages exchanged between healthcare applications are ultimately recorded into local healthcare databases, mostly in relational databases. In order to bring these relational databases in compliance with HL7, mappings between HL7 RIM (Reference Information Model) and relational database schema are required. Currently, RIM and database mapping is largely performed manually, therefore it is tedious, time consuming, error prone and expensive process. It is a challenging task to determine all correspondences between RIM and schema automatically because of extreme heterogeneity issues in healthcare databases. To reduce the amount of manual efforts as much as possible, autonomous mapping approaches are required. This paper proposes a technique that addresses the aforementioned mapping issue and aligns healthcare databases to HL7 V3 RIM specifications. Furthermore, the proposed technique has been implemented as a working application and tested on real world healthcare systems. The application loads the target healthcare schema and then identifies the most appropriate match for tables and the associated fields in the schema by using domain knowledge and the matching rules defined in the Mapping Knowledge Repository. These rules are designed to handle the complexity of semantics found in healthcare databases. The GUI allows users to view and edit/re-map the correspondences. Once all the mappings are defined, the application generates Mapping Specification, which contains all the mapping information i.e. database tables and fields with associated RIM classes and attributes. In order to enable the transactions, the application is facilitated with the autonomous code generation from the Mapping Specification. The Code Generator component focuses primarily on generating custom classes and hibernate mapping files against the runtime system to retrieve and parse the data from the data source—thus allows bi-directional HL7 to database communication, with minimum programming required. Our experimental results show 35–65% accuracy on real laboratory systems, thus demonstrating the promise of the approach. The proposed scheme is an effective step in bringing the clinical databases in compliance with RIM, providing ease and flexibility.     

A localization algorithm for large scale mobile wireless sensor networks: a learning approach

Localization is a crucial problem in wireless sensor networks and most of the localization algorithms given in the literature are non-adaptive and designed for fixed sensor networks. In this paper, we propose a learning based localization algorithm for mobile wireless sensor networks. By this technique, mobility in the network will be discovered by two crucial methods in the beacons: position and distance checks methods. These two methods help to have accurate localization and constrain communication just when it is necessary. The proposed method localizes the nodes based on connectivity information (hop count), which doesn’t need extra hardware and is cost efficient. The experimental results show that the proposed algorithm is scalable with a small set of beacons in large scale network with a high density of nodes. The given algorithm is fast and free from a pre-deployment requirement. The simulation results show the high performance of the proposed algorithm.     

Speaker recognition utilizing distributed DCT-II based Mel frequency cepstral coefficients and fuzzy vector quantization

In this paper, a new and novel Automatic Speaker Recognition (ASR) system is presented. The new ASR system includes novel feature extraction and vector classification steps utilizing distributed Discrete Cosine Transform (DCT-II) based Mel Frequency Cepstral Coefficients (MFCC) and Fuzzy Vector Quantization (FVQ). The ASR algorithm utilizes an approach based on MFCC to identify dynamic features that are used for Speaker Recognition (SR). A series of experiments were performed utilizing three different feature extraction methods: (1) conventional MFCC; (2) Delta-Delta MFCC (DDMFCC); and (3) DCT-II based DDMFCC. The experiments were then expanded to include four classifiers: (1) FVQ; (2) K-means Vector Quantization (VQ); (3) Linde, Buzo and Gray VQ; and (4) Gaussian Mixed Model (GMM). The combination of DCT-II based MFCC, DMFCC and DDMFCC with FVQ was found to have the lowest Equal Error Rate for the VQ based classifiers. The results found were an improvement over previously reported non-GMM methods and approached the results achieved for the computationally expensive GMM based method. Speaker verification tests carried out highlighted the overall performance improvement for the new ASR system. The National Institute of Standards and Technology Speaker Recognition Evaluation corpora was used to provide speaker source data for the experiments.     

Yet another cybersecurity risk assessment framework

IT systems pervade our society more and more, and we become heavily dependent on them. At the same time, these systems are increasingly targeted in cyberattacks, making us vulnerable. Enterprise and cybersecurity responsibles face the problem of defining techniques that raise the level of security. They need to decide which mechanism provides the most efficient defense with limited resources. Basically, the risks need to be assessed to determine the best cost-to-benefit ratio. One way to achieve this is through threat modeling; however, threat modeling is not commonly used in the enterprise IT risk domain. Furthermore, the existing threat modeling methods have shortcomings. This paper introduces a metamodel-based approach named Yet Another Cybersecurity Risk Assessment Framework (Yacraf). Yacraf aims to enable comprehensive risk assessment for organizations with more decision support. The paper includes a risk calculation formalization and also an example showing how an organization can use and benefit from Yacraf.

     

水声网络全双工定向碰撞避免媒体接入控制协议

近年来水声网络(UAN)技术飞速发展,但仍然面临诸多严峻挑战,能量效率成为水声网络的首要考虑因素。此外,水声信道传播时延大且可用带宽受限,严重制约了水声通信技术的可靠性和有效性,进而限制了水声网络的整体性能。定向通信技术可以有效改善上述情况,通过波束聚焦能力将声波能量聚集在一定角度范围内,获得比全向通信更高的通信范围和信噪比,提升整个网络的能耗效率和空间复用率。但该技术需要对目的节点的位置具备先验知识,且会面临“聋节点”问题,因此该文提出一个水声网络全双工定向碰撞避免(FDDCA)媒体接入控制(MAC)协议,通过装备两个分别处于不同工作频带的全向换能器和定向换能器解决“聋节点”问题,并通过降低节点的冲突域解决了“暴露终端”问题。仿真结果表明,与水下Aloha(UW-Aloha)和时隙地面多址接入(S-FAMA)协议相比,FDDCA在多汇聚节点的网络拓扑下吞吐量分别提升了140%和400%,网络能效上节省了90%和94%。     

Theoretical Investigation of Cubic BaVO<Subscript>3</Subscript> and LaVO<Subscript>3</Subscript> Perovskites via Tran–Blaha-Modified Becke–Johnson Exchange Potential Approach

The full potential linearized augmented plane wave method of density functional theory has been used to investigate the structural, electronic, magnetic and thermoelectric properties of cubic perovskites BaVO₃ and LaVO₃. The ferromagnetic ground state has been found to be stable by comparing the total energies of non-spin-polarized and spin-polarized calculations performed for optimized unit cells. For both compounds, the bond length and tolerance factor are also measured. From the band structures and density of states plots, it is found that both compounds are half-metallic. We found that the presence of V at the octahedral site of these perovskites develops exchange splitting through p-d hybridization, which results in a stable ferromagnetic state. The observed exchange splitting is further clarified from the magnetic moment, charge and spin of the anion and cations. Finally, we also presented the calculated thermoelectric properties of these materials, which show that half-metallic BaVO₃ and LaVO₃ materials are potential contenders for thermoelectric applications.     

Optimization of a proton exchange membrane fuel cell membrane electrode assembly

A computational framework for fuel cell analysis and optimization is presented as an innovative alternative to the time consuming trial-and-error process currently used for fuel cell design. The framework is based on a two-dimensional through-the-channel isothermal, isobaric and single phase membrane electrode assembly (MEA) model. The model input parameters are the manufacturing parameters used to build the MEA: platinum loading, platinum to carbon ratio, electrolyte content and gas diffusion layer porosity. The governing equations of the fuel cell model are solved using Netwon’s algorithm and an adaptive finite element method in order to achieve near quadratic convergence and a mesh independent solution respectively. The analysis module is used to solve the optimization problem of finding the optimal MEA composition for maximizing performance. To solve the optimization problem a gradient-based optimization algorithm is used in conjunction with analytical sensitivities. By using a gradient-based method and analytical sensitivities, the framework presented is capable of solving a complete MEA optimization problem with state-of-the-art electrode models in approximately 30 min, making it a viable alternative for solving large-scale fuel cell problems.     

Learning from demonstration in robots: Experimental comparison of neural architectures

Robots have played an important role in the automation of computer aided manufacturing. The classical robot control implementation involves an expensive key step of model-based programming. An intuitive way to reduce this expensive exercise is to replace programming with machine learning of robot actions from demonstration where a (learner) robot learns an action by observing a demonstrator robot performing the same. To achieve this learning from demonstration (LFD) different machine learning techniques such as Artificial Neural Networks (ANN), Genetic Algorithms, Hidden Markov Models, Support Vector Machines, etc. can be used. This piece of work focuses exclusively on ANNs. Since ANNs have many standard architectural variations divided into two basic computational categories namely the recurrent networks and feed-forward networks, representative networks from each have been selected for study, i.e. Feed Forward Multilayer Perceptron (FF) network for feed-forward networks category and Elman (EL), and Nonlinear Autoregressive Exogenous Model (NARX) networks for the recurrent networks category. The main objective of this work is to identify the most suitable neural architecture for application of LFD in learning different robot actions. The sensor and actuator streams of demonstrated action are used as training data for ANN learning. Consequently, the learning capability is measured by comparing the error between demonstrator and corresponding learner streams. To achieve fairness in comparison three steps have been taken. First, Dynamic Time Warping is used to measure the error between demonstrator and learner streams, which gives resilience against translation in time. Second, comparison statistics are drawn between the best, instead of weight-equal, configurations of competing architectures so that learning capability of any architecture is not forced handicap. Third, each configuration's error is calculated as the average of ten trials of all possible learning sequences with random weight initialization so that the error value is independent of a particular sequence of learning or a particular set of initial weights. Six experiments are conducted to get a performance pattern of each architecture. In each experiment, a total of nine different robot actions were tested. Error statistics thus obtained have shown that NARX architecture is most suitable for this learning problem whereas Elman architecture has shown the worst suitability. Interestingly the computationally lesser MLP gives much lower and slightly higher error statistics compared to the computationally superior Elman and NARX neural architectures, respectively.