Feasibility of diffusion kurtosis imaging in evaluating cervical spinal cord injury in multiple sclerosis.

This research aimed to assess gray matter (GM), white matter (WM), lesions of multiple sclerosis (MS) and the therapeutic effect using diffusion kurtosis imaging (DKI). From January 2018 to October 2019, 78 subjects (48 of MS and 30 of health) perform routine MR scan and DKI of cervical spinal cord. The MS patients were divided into 2 groups according to the presence or absence of T2 hyperintensity. DKI-metrics were measured in the lesions, normal-appearing GM and WM. Significant differences were detected in DKI metrics between MS and healthy (P < .05) and between patients with cervical spinal cord T2-hyperintense and without T2-hyperintense (P < .001). Compared to healthy, GM-mean kurtosis (MK), GM-radial kurtosis, and WM-fractional anisotropy, WM-axial diffusion were statistically reduced in patients without T2-hyperintense (P < .05). Significant differences were observed in DKI metrics between patients with T2-hyperintense after therapy (P < .05), as well as GM-MK and WM-fractional anisotropy, WM-axial diffusion in patients without T2-hyperintense (P < .05); Expanded Disability Status Scale was correlated with MK values, as well as Expanded Disability Status Scale scores and MK values after therapy. Our results indicate that DKI-metrics can detect and quantitatively evaluate the changes in cervical spinal cord micropathological structure.

Non-contrast CT-based radiomic signature of pericoronary adipose tissue for screening non-calcified plaque.

Objective.To develop two combined clinical-radiomics models of pericoronary adipose tissue (PCAT) for the presence and characterization of non-calcified plaques on non-contrast CT scan.Approach.Altogether, 431 patients undergoing Coronary Computed Tomography Angiography from March 2019 to June 2021 who had complete data were enrolled, including 173 patients with non-calcified plaques of the right coronary artery(RCA) and 258 with no abnormality. PCAT was segmented around the proximal RCA on non-contrast CT scan (calcium score acquisition). Two best models were established by screening features and classifiers respectively using FeAture Explorer software. Model 1 distinguished normal coronary arteries from those with non-calcified plaques, and model 2 distinguished vulnerable plaques in non-calcified plaques. Performance was assessed by the area under the receiver operating characteristic curve (AUC-ROC).Main results.4 and 9 features were selected for the establishment of the radiomics model respectively through Model 1 and 2. In the test group, the AUC values, sensitivity, specificity and accuracy were 0.833%, 78.3%, 80.8%, 76.6% and 0.7467%, 75.0%, 77.8%, 73.5%, respectively. The combined model including radiomics features and independent clinical factors yielded an AUC, sensitivity, specificity and accuracy of 0.896%, 81.4%, 86.5%, 77.9% for model 1 and 0.752%, 75.0%, 77.8%, 73.5% for model 2.Significance.The combined clinical-radiomics models based on non-contrast CT images of PCAT had good diagnostic efficacy for non-calcified and vulnerable plaques.

Evaluation of microstructural changes in spinal cord of patients with degenerative cervical myelopathy by diffusion kurtosis imaging and investigate the correlation with JOA score.

BACKGROUND: To explore the feasibility of the metrics of diffusion kurtosis imaging (DKI) for investigations of the microstructural changes of spinal cord injury in patients with degenerative cervical myelopathy (DCM) and the correlation between Japan Orthopaedic Association (JOA) scores and DKI metrics. METHODS: Fifty-seven patients with DCM and 38 healthy volunteers underwent 3.0 T magnetic resonance (MR) imaging with routine MRI sequences and DKI from echo-planar imaging sequence. Based on the JOA score, DCM patients were divided into four subgroups. DKI metrics of the DCM group and control group were obtained and compared, separately for the white matter (WM) and the gray matter (GM). RESULTS: The FA values in WM were significantly lower (P = 0.020) in the DCM group than in the control group. The MK values in GM were lower (P = 0.011) in the DCM group than in the control group. The MD values in WM were significantly higher (P = 0.010) in the DCM group than in the control group. In GM, the JOA score was positively correlated with the MK values (r = 0.768, P < 0.05). In the WM, the JOA score was positively correlated with the FA values (r = 0.612, P < 0.05). CONCLUSION: DKI provides quantitive evaluation to the characters of microstructure of the spinal cord damage in patients with DCM compared to conventional MR. MK values can reflect microstructural abnormalities of gray matter of the cervical spinal cord and provide more information beyond that obtained with routine diffusion metrics. In addition, MK values of GM and FA values of WM may as a be highly sensitive biomarker for the degree of cervical spinal cord damage.

A DAG scheduling scheme on heterogeneous computing systems using tuple-based chemical reaction optimization.

A complex computing problem can be solved efficiently on a system with multiple computing nodes by dividing its implementation code into several parallel processing modules or tasks that can be formulated as directed acyclic graph (DAG) problems. The DAG jobs may be mapped to and scheduled on the computing nodes to minimize the total execution time. Searching an optimal DAG scheduling solution is considered to be NP-complete. This paper proposed a tuple molecular structure-based chemical reaction optimization (TMSCRO) method for DAG scheduling on heterogeneous computing systems, based on a very recently proposed metaheuristic method, chemical reaction optimization (CRO). Comparing with other CRO-based algorithms for DAG scheduling, the design of tuple reaction molecular structure and four elementary reaction operators of TMSCRO is more reasonable. TMSCRO also applies the concept of constrained critical paths (CCPs), constrained-critical-path directed acyclic graph (CCPDAG) and super molecule for accelerating convergence. In this paper, we have also conducted simulation experiments to verify the effectiveness and efficiency of TMSCRO upon a large set of randomly generated graphs and the graphs for real world problems.

Web service reputation evaluation based on QoS measurement.

In the early service transactions, quality of service (QoS) information was published by service provider which was not always true and credible. For better verification the trust of the QoS information was provided by the Web service. In this paper, the factual QoS running data are collected by our WS-QoS measurement tool; based on these objectivity data, an algorithm compares the difference of the offered and measured quality data of the service and gives the similarity, and then a reputation evaluation method computes the reputation level of the Web service based on the similarity. The initial implementation and experiment with three Web services' example show that this approach is feasible and these values can act as the references for subsequent consumers to select the service.

Novel web service selection model based on discrete group search.

In our earlier work, we present a novel formal method for the semiautomatic verification of specifications and for describing web service composition components by using abstract concepts. After verification, the instantiations of components were selected to satisfy the complex service performance constraints. However, selecting an optimal instantiation, which comprises different candidate services for each generic service, from a large number of instantiations is difficult. Therefore, we present a new evolutionary approach on the basis of the discrete group search service (D-GSS) model. With regard to obtaining the optimal multiconstraint instantiation of the complex component, the D-GSS model has competitive performance compared with other service selection models in terms of accuracy, efficiency, and ability to solve high-dimensional service composition component problems. We propose the cost function and the discrete group search optimizer (D-GSO) algorithm and study the convergence of the D-GSS model through verification and test cases.

Improving causality induction with category learning.

Causal relations are of fundamental importance for human perception and reasoning. According to the nature of causality, causality has explicit and implicit forms. In the case of explicit form, causal-effect relations exist at either clausal or discourse levels. The implicit causal-effect relations heavily rely on empirical analysis and evidence accumulation. This paper proposes a comprehensive causality extraction system (CL-CIS) integrated with the means of category-learning. CL-CIS considers cause-effect relations in both explicit and implicit forms and especially practices the relation between category and causality in computation. In elaborately designed experiments, CL-CIS is evaluated together with general causality analysis system (GCAS) and general causality analysis system with learning (GCAS-L), and it testified to its own capability and performance in construction of cause-effect relations. This paper confirms the expectation that the precision and coverage of causality induction can be remarkably improved by means of causal and category learning.

Self-supervised Chinese ontology learning from online encyclopedias.

Constructing ontology manually is a time-consuming, error-prone, and tedious task. We present SSCO, a self-supervised learning based chinese ontology, which contains about 255 thousand concepts, 5 million entities, and 40 million facts. We explore the three largest online Chinese encyclopedias for ontology learning and describe how to transfer the structured knowledge in encyclopedias, including article titles, category labels, redirection pages, taxonomy systems, and InfoBox modules, into ontological form. In order to avoid the errors in encyclopedias and enrich the learnt ontology, we also apply some machine learning based methods. First, we proof that the self-supervised machine learning method is practicable in Chinese relation extraction (at least for synonymy and hyponymy) statistically and experimentally and train some self-supervised models (SVMs and CRFs) for synonymy extraction, concept-subconcept relation extraction, and concept-instance relation extraction; the advantages of our methods are that all training examples are automatically generated from the structural information of encyclopedias and a few general heuristic rules. Finally, we evaluate SSCO in two aspects, scale and precision; manual evaluation results show that the ontology has excellent precision, and high coverage is concluded by comparing SSCO with other famous ontologies and knowledge bases; the experiment results also indicate that the self-supervised models obviously enrich SSCO.