Numerical assessment of stability of interface discontinuous finite element pressure spaces

The stability of two recently developed pressure spaces has been assessed numerically: The space proposed by Ausas et al. [R.F. Ausas, F.S. Sousa, G.C. Buscaglia, An improved finite element space for discontinuous pressures, Comput. Methods Appl. Mech. Engrg. 199 (2010) 1019-1031], which is capable of representing discontinuous pressures, and the space proposed by Coppola-Owen and Codina [A.H. Coppola-Owen, R. Codina, Improving Eulerian two-phase flow finite element approximation with discontinuous gradient pressure shape functions, Int. J. Numer. Methods Fluids, 49 (2005) 1287-1304], which can represent discontinuities in pressure gradients. We assess the stability of these spaces by numerically computing the inf-sup constants of several meshes. The inf-sup constant results as the solution of a generalized eigenvalue problems. Both spaces are in this way confirmed to be stable in their original form. An application of the same numerical assessment tool to the stabilized equal-order P-1/P-1 formulation is then reported. An interesting finding is that the stabilization coefficient can be safely set to zero in an arbitrary band of elements without compromising the formulation's stability. An analogous result is also reported for the mini-element P-1(+)/P-1 when the velocity bubbles are removed in an arbitrary band of elements. (C) 2012 Elsevier B.V. All rights reserved.

Experiência inicial com a técnica de Pomerantzeff para redução do tamanho do átrio esquerdo gigante

INTRODUÇÃO: A mais comum indicação de correção cirúrgica de átrio esquerdo gigante está associada à insuficiência da valva mitral, com ou sem fibrilação atrial. Diversas técnicas para este fim já estão descritas com resultados variáveis. OBJETIVO: Apresentar a experiência inicial com a técnica da ressecção triangular tangencial (Pomerantzeff). MÉTODOS: De 2002 a 2010, quatro pacientes foram submetidos a operação da valva mitral com redução do volume do átrio esquerdo pela técnica da ressecção triangular tangencial em nosso serviço. Três pacientes eram do sexo feminino. A idade variou de 21 a 51 anos. Os quatro pacientes encontravam-se com fibrilação atrial. A fração de ejeção do ventrículo esquerdo no pré-operatório variava de 38% a 62%. O diâmetro do átrio esquerdo variou de 78 a 140 mm. Após o tratamento da disfunção mitral, o átrio esquerdo foi reduzido por meio de ressecção triangular tangencial da sua parede posterior, entre as veias pulmonares, para evitar distorções anatômicas do anel mitral ou veias pulmonares, reduzindo a tensão na linha de sutura. RESULTADOS: Tempo médio de internação hospitalar foi de 21,5 ± 6,5 dias. O tempo de circulação extracorpórea médio foi de 130 ± 30 minutos. Não houve sangramento cirúrgico ou mortalidade no período pós-operatório. Todos os pacientes tiveram o ritmo sinusal restabelecido na saída de circulação extracorpórea, mantendo esse ritmo no pós-operatório. O diâmetro médio do átrio esquerdo foi reduzido em 50,5 ± 19,5%. A fração de ejeção do ventrículo esquerdo melhorou em todas as pacientes. CONCLUSÃO: Os resultados iniciais com essa técnica têm demonstrado redução efetiva do átrio esquerdo.

CID: an efficient complexity-invariant distance for time series

The ubiquity of time series data across almost all human endeavors has produced a great interest in time series data mining in the last decade. While dozens of classification algorithms have been applied to time series, recent empirical evidence strongly suggests that simple nearest neighbor classification is exceptionally difficult to beat. The choice of distance measure used by the nearest neighbor algorithm is important, and depends on the invariances required by the domain. For example, motion capture data typically requires invariance to warping, and cardiology data requires invariance to the baseline (the mean value). Similarly, recent work suggests that for time series clustering, the choice of clustering algorithm is much less important than the choice of distance measure used.In this work we make a somewhat surprising claim. There is an invariance that the community seems to have missed, complexity invariance. Intuitively, the problem is that in many domains the different classes may have different complexities, and pairs of complex objects, even those which subjectively may seem very similar to the human eye, tend to be further apart under current distance measures than pairs of simple objects. This fact introduces errors in nearest neighbor classification, where some complex objects may be incorrectly assigned to a simpler class. Similarly, for clustering this effect can introduce errors by “suggesting” to the clustering algorithm that subjectively similar, but complex objects belong in a sparser and larger diameter cluster than is truly warranted.We introduce the first complexity-invariant distance measure for time series, and show that it generally produces significant improvements in classification and clustering accuracy. We further show that this improvement does not compromise efficiency, since we can lower bound the measure and use a modification of triangular inequality, thus making use of most existing indexing and data mining algorithms. We evaluate our ideas with the largest and most comprehensive set of time series mining experiments ever attempted in a single work, and show that complexity-invariant distance measures can produce improvements in classification and clustering in the vast majority of cases.