BM+-tree: A hyperplane-based index method for high-dimensional metric spaces

In this paper, we propose a novel high-dimensional index method, the BM+-tree, to support efficient processing of similarity search queries in high-dimensional spaces. The main idea of the proposed index is to improve data partitioning efficiency in a high-dimensional space by using a rotary binary hyperplane, which further partitions a subspace and can also take advantage of the twin node concept used in the M+-tree. Compared with the key dimension concept in the M+-tree, the binary hyperplane is more effective in data filtering. High space utilization is achieved by dynamically performing data reallocation between twin nodes. In addition, a post processing step is used after index building to ensure effective filtration. Experimental results using two types of real data sets illustrate a significantly improved filtering efficiency.

Compiling a caseload index for mental health case management

Objective: The purpose of this study was to explore methods of determining an appropriate caseload for mental health case managers. Method: Seven factors that may impinge on case manager performance and impact on caseload were identified, having reference to published literature and service practice in Victoria and Queensland. The advantages and disadvantages of including these factors in a caseload index were evaluated. Results: Three caseload index methodologies are presented. Each method makes use of different data and has advantages and disadvantages. There is a trade-off between simplicity and ease of application and the comprehensive use of relevant information. Methods vary in their implications for service efficiency and equity in workload distribution. Conclusions: Caseload is a key issue in service planning and staff management. Factors that have the potential to contribute to caseload can be readily identified. However, there is likely to be disagreement as to the weight assigned to any factor and the approach taken may depend on the purpose and context of the caseload calculation. A great deal more research is required to provide an empirical basis for algorithms used in caseload calculation.

Magnetic resonance imaging analysis of the upper cervical spine extensor musculature in an asymptomatic cohort: an index of fat within muscle

AIM: To establish a simple method to quantify muscle/fat constituents in cervical muscles of asymptomatic women using magnetic resonance imaging (MRI), and to determine whether there is an age effect within a defined age range. MATERIALS AND METHODS: MRI of the upper cervical spine was performed for 42 asymptomatic women aged 18-45 years. The muscle and fat signal intensities on axial spin echo T1-weighted images were quantitatively classified by taking a ratio of the pixel intensity profiles of muscle against those of intermuscular fat for the rectus capitis posterior major and minor and inferior obliquus capitis muscles bilaterally. Inter- and intra-examiner agreement was scrutinized. RESULTS: The average relative values of fat within the upper cervical musculature compared with intermuscular fat indicated that there were only slight variations in indices between the three sets of muscles. There was no significant correlation between age and fat indices. There were significant differences for the relative fat within the muscle compared with intermuscular fat and body mass index for the right rectus capitis posterior major and right and left inferior obliquus capitis muscles (p = 0.032). Intraclass correlation coefficients for intraobserver agreement ranged from 0.94 to 0.98. Inter-rater agreement of the measurements ranged from 0.75 to 0.97. CONCLUSION: A quantitative measure of muscle/fat constituents has been developed, and results of this study indicate that relative fatty infiltration is not a feature of age in the upper cervical extensor muscles of women aged 18-45 years. (C) 2005 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.

Multiple imputation for body mass index: lessons from the Australian Longitudinal Study on Women's Health

In large epidemiological studies missing data can be a problem, especially if information is sought on a sensitive topic or when a composite measure is calculated from several variables each affected by missing values. Multiple imputation is the method of choice for 'filling in' missing data based on associations among variables. Using an example about body mass index from the Australian Longitudinal Study on Women's Health, we identify a subset of variables that are particularly useful for imputing values for the target variables. Then we illustrate two uses of multiple imputation. The first is to examine and correct for bias when data are not missing completely at random. The second is to impute missing values for an important covariate; in this case omission from the imputation process of variables to be used in the analysis may introduce bias. We conclude with several recommendations for handling issues of missing data. Copyright (C) 2004 John Wiley Sons, Ltd.

Validation and patient acceptance of a computer touch screen version of the WOMAC 3.1 osteoarthritis index

Objectives: To validate the WOMAC 3.1 in a touch screen computer format, which applies each question as a cartoon in writing and in speech (QUALITOUCH method), and to assess patient acceptance of the computer touch screen version. Methods: The paper and computer formats of WOMAC 3.1 were applied in random order to 53 subjects with hip or knee osteoarthritis. The mean age of the subjects was 64 years ( range 45 to 83), 60% were male, 53% were 65 years or older, and 53% used computers at home or at work. Agreement between formats was assessed by intraclass correlation coefficients (ICCs). Preferences were assessed with a supplementary questionnaire. Results: ICCs between formats were 0.92 (95% confidence interval, 0.87 to 0.96) for pain; 0.94 (0.90 to 0.97) for stiffness, and 0.96 ( 0.94 to 0.98) for function. ICCs were similar in men and women, in subjects with or without previous computer experience, and in subjects below or above age 65. The computer format was found easier to use by 26% of the subjects, the paper format by 8%, and 66% were undecided. Overall, 53% of subjects preferred the computer format, while 9% preferred the paper format, and 38% were undecided. Conclusion: The computer format of the WOMAC 3.1 is a reliable assessment tool. Agreement between computer and paper formats was independent of computer experience, age, or sex. Thus the computer format may help improve patient follow up by meeting patients' preferences and providing immediate results.

On the equivalence of weighted Country-Product-Dummy (CPD) Method and the Rao-System for multilateral price comparisons

The country-product-dummy (CPD) method, originally proposed in Summers (1973), has recently been revisited in its weighted formulation to handle a variety of data related situations (Rao and Timmer, 2000, 2003; Heravi et al., 2001; Rao, 2001; Aten and Menezes, 2002; Heston and Aten, 2002; Deaton et al., 2004). The CPD method is also increasingly being used in the context of hedonic modelling instead of its original purpose of filling holes in Summers (1973). However, the CPD method is seen, among practitioners, as a black box due to its regression formulation. The main objective of the paper is to establish equivalence of purchasing power parities and international prices derived from the application of the weighted-CPD method with those arising out of the Rao-system for multilateral comparisons. A major implication of this result is that the weighted-CPD method would then be a natural method of aggregation at all levels of aggregation within the context of international comparisons.

An adaptive and dynamic dimensionality reduction method for high-dimensional indexing

The notorious "dimensionality curse" is a well-known phenomenon for any multi-dimensional indexes attempting to scale up to high dimensions. One well-known approach to overcome degradation in performance with respect to increasing dimensions is to reduce the dimensionality of the original dataset before constructing the index. However, identifying the correlation among the dimensions and effectively reducing them are challenging tasks. In this paper, we present an adaptive Multi-level Mahalanobis-based Dimensionality Reduction (MMDR) technique for high-dimensional indexing. Our MMDR technique has four notable features compared to existing methods. First, it discovers elliptical clusters for more effective dimensionality reduction by using only the low-dimensional subspaces. Second, data points in the different axis systems are indexed using a single B+-tree. Third, our technique is highly scalable in terms of data size and dimension. Finally, it is also dynamic and adaptive to insertions. An extensive performance study was conducted using both real and synthetic datasets, and the results show that our technique not only achieves higher precision, but also enables queries to be processed efficiently. Copyright Springer-Verlag 2005

Ab initio studies of hydrogen desorption from low index magnesium hydride surface

The low index Magnesium hydride surfaces, MgH2(001) and MgH2(110), have been studied by ab intio Density Functional Theory (DFT) calculations. It was found that the MgH2(110) surface is more stable than MgH2(001) surface, which is in good agreement with the experimental observation. The H-2 desorption barriers vary depending on the crystalline surfaces that are exposed and also the specific H atom sites involved-they are found to be generally high, due to the thermodynamic stability of the MgH2, system, and are larger for the MgH2(001) surface. The pathway for recombinative desorption of one in-plane and one bridging H atom from the MgH2(110) surface was found to be the lowest energy barrier amongst those computed (172 KJ/mol) and is in good agreement with the experimental estimates. (c) 2006 Elsevier B.V. All rights reserved.

Use of drought response index for identification of drought tolerant genotypes in rainfed lowland rice

Drought is a major constraint for rice production in the rainfed lowlands in Southeast Asia and Eastern India. The breeding programs for tainted lowland rice in these regions focus on adaptation to a range of drought conditions. However, a method of selection of drought tolerant genotypes has not been established and is considered to be one of the constraints faced by rice breeders. Drought response index (DRI) is based on grain yield adjusted for variation in potential yield and flowering date, and has been used recently, but its consistency among drought environments and hence its usefulness is not certain. In order to establish a selection method and subsequently to identify donor parents for drought resistance breeding, a series of experiments with 15 contrasting genotypes was conducted under well-watered and managed drought conditions at two sites for 5 years in Cambodia. Water level in the field was recorded and used to estimate the relative water level (WLREL) around flowering as an index of the severity of water deficit at the time of flowering for each entry. This was used to determine if DRI or yield reduction was due to drought tolerance or related to the amount of available water at flowering, i.e. drought escape. Grain yield reduction due to drought ranged from 12 to 46%. The drought occurred mainly during the reproductive phase, while four experiments had water stress from the early vegetative stage. There was significant variation for water availability around flowering among the nine experiments and this was associated with variation in mean yield reduction. Genotypic variation in DRI was consistent among most experiments, and genotypic mean DRI ranged from -0.54 to 0.47 (LSD 5% = 0.47). Genotypic variation in DRI was not related to WLREL around flowering in the nine environments. It is concluded that selection for DRI under drought conditions would allow breeders to identify donor lines with high drought tolerance as an important component of breeding better adapted varieties for the rainfed lowlands; two genotypes were identified with high DRI and low yield reduction and were subsequently used in the breeding program in Cambodia. (c) 2006 Elsevier B.V. All rights reserved.

A new indexing method for high dimensional dataset

Indexing high dimensional datasets has attracted extensive attention from many researchers in the last decade. Since R-tree type of index structures are known as suffering curse of dimensionality problems, Pyramid-tree type of index structures, which are based on the B-tree, have been proposed to break the curse of dimensionality. However, for high dimensional data, the number of pyramids is often insufficient to discriminate data points when the number of dimensions is high. Its effectiveness degrades dramatically with the increase of dimensionality. In this paper, we focus on one particular issue of curse of dimensionality; that is, the surface of a hypercube in a high dimensional space approaches 100% of the total hypercube volume when the number of dimensions approaches infinite. We propose a new indexing method based on the surface of dimensionality. We prove that the Pyramid tree technology is a special case of our method. The results of our experiments demonstrate clear priority of our novel method.