Application of data mining techniques in the prediction of coronary artery disease : use of anaesthesia time-series and patient risk factor data

The high morbidity and mortality associated with atherosclerotic coronary vascular disease (CVD) and its complications are being lessened by the increased knowledge of risk factors, effective preventative measures and proven therapeutic interventions. However, significant CVD morbidity remains and sudden cardiac death continues to be a presenting feature for some subsequently diagnosed with CVD. Coronary vascular disease is also the leading cause of anaesthesia related complications. Stress electrocardiography/exercise testing is predictive of 10 year risk of CVD events and the cardiovascular variables used to score this test are monitored peri-operatively. Similar physiological time-series datasets are being subjected to data mining methods for the prediction of medical diagnoses and outcomes. This study aims to find predictors of CVD using anaesthesia time-series data and patient risk factor data. Several pre-processing and predictive data mining methods are applied to this data. Physiological time-series data related to anaesthetic procedures are subjected to pre-processing methods for removal of outliers, calculation of moving averages as well as data summarisation and data abstraction methods. Feature selection methods of both wrapper and filter types are applied to derived physiological time-series variable sets alone and to the same variables combined with risk factor variables. The ability of these methods to identify subsets of highly correlated but non-redundant variables is assessed. The major dataset is derived from the entire anaesthesia population and subsets of this population are considered to be at increased anaesthesia risk based on their need for more intensive monitoring (invasive haemodynamic monitoring and additional ECG leads). Because of the unbalanced class distribution in the data, majority class under-sampling and Kappa statistic together with misclassification rate and area under the ROC curve (AUC) are used for evaluation of models generated using different prediction algorithms. The performance based on models derived from feature reduced datasets reveal the filter method, Cfs subset evaluation, to be most consistently effective although Consistency derived subsets tended to slightly increased accuracy but markedly increased complexity. The use of misclassification rate (MR) for model performance evaluation is influenced by class distribution. This could be eliminated by consideration of the AUC or Kappa statistic as well by evaluation of subsets with under-sampled majority class. The noise and outlier removal pre-processing methods produced models with MR ranging from 10.69 to 12.62 with the lowest value being for data from which both outliers and noise were removed (MR 10.69). For the raw time-series dataset, MR is 12.34. Feature selection results in reduction in MR to 9.8 to 10.16 with time segmented summary data (dataset F) MR being 9.8 and raw time-series summary data (dataset A) being 9.92. However, for all time-series only based datasets, the complexity is high. For most pre-processing methods, Cfs could identify a subset of correlated and non-redundant variables from the time-series alone datasets but models derived from these subsets are of one leaf only. MR values are consistent with class distribution in the subset folds evaluated in the n-cross validation method. For models based on Cfs selected time-series derived and risk factor (RF) variables, the MR ranges from 8.83 to 10.36 with dataset RF_A (raw time-series data and RF) being 8.85 and dataset RF_F (time segmented time-series variables and RF) being 9.09. The models based on counts of outliers and counts of data points outside normal range (Dataset RF_E) and derived variables based on time series transformed using Symbolic Aggregate Approximation (SAX) with associated time-series pattern cluster membership (Dataset RF_ G) perform the least well with MR of 10.25 and 10.36 respectively. For coronary vascular disease prediction, nearest neighbour (NNge) and the support vector machine based method, SMO, have the highest MR of 10.1 and 10.28 while logistic regression (LR) and the decision tree (DT) method, J48, have MR of 8.85 and 9.0 respectively. DT rules are most comprehensible and clinically relevant. The predictive accuracy increase achieved by addition of risk factor variables to time-series variable based models is significant. The addition of time-series derived variables to models based on risk factor variables alone is associated with a trend to improved performance. Data mining of feature reduced, anaesthesia time-series variables together with risk factor variables can produce compact and moderately accurate models able to predict coronary vascular disease. Decision tree analysis of time-series data combined with risk factor variables yields rules which are more accurate than models based on time-series data alone. The limited additional value provided by electrocardiographic variables when compared to use of risk factors alone is similar to recent suggestions that exercise electrocardiography (exECG) under standardised conditions has limited additional diagnostic value over risk factor analysis and symptom pattern. The effect of the pre-processing used in this study had limited effect when time-series variables and risk factor variables are used as model input. In the absence of risk factor input, the use of time-series variables after outlier removal and time series variables based on physiological variable values’ being outside the accepted normal range is associated with some improvement in model performance.

Difference in mosquito species (Diptera: Culicidae) and the transmission of Ross River Virus between coastline and inland areas in Brisbane, Australia

This study examined the distribution of major mosquito species and their roles in the transmission of Ross River virus (RRV) infection for coastline and inland areas in Brisbane, Australia (27°28′ S, 153°2′ E). We obtained data on the monthly counts of RRV cases in Brisbane between November 1998 and December 2001 by statistical local areas from the Queensland Department of Health and the monthly mosquito abundance from the Brisbane City Council. Correlation analysis was used to assess the pairwise relationships between mosquito density and the incidence of RRV disease. This study showed that the mosquito abundance of Aedes vigilax (Skuse), Culex annulirostris (Skuse), and Aedes vittiger (Skuse) were significantly associated with the monthly incidence of RRV in the coastline area, whereas Aedes vigilax, Culex annulirostris, and Aedes notoscriptus (Skuse) were significantly associated with the monthly incidence of RRV in the inland area. The results of the classification and regression tree (CART) analysis show that both occurrence and incidence of RRV were influenced by interactions between species in both coastal and inland regions. We found that there was an 89% chance for an occurrence of RRV if the abundance of Ae. vigifax was between 64 and 90 in the coastline region. There was an 80% chance for an occurrence of RRV if the density of Cx. annulirostris was between 53 and 74 in the inland area. The results of this study may have applications as a decision support tool in planning disease control of RRV and other mosquito-borne diseases.

Importance of soil ecology in vegetation rehabilitation on a North Stradbroke Island sand mine

Sibelco Australia Limited (SAL), a mineral sand mining operation on North Stradbroke Island, undertakes progressive rehabilitation of mined areas. Initial investigations have found that some areas at SAL’s Yarraman Mine have failed to redevelop towards approved criteria. This study, undertaken in 2010, examined ground cover rehabilitation of different aged plots at the Yarraman Mine to determine if there was a relationship between key soil and vegetation attributes. Vegetation and soil data were collected from five plots rehabilitated in 2003, 2006, 2008, 2009 and 2010, and one unmined plot. Cluster (PATN) analysis revealed that vegetation species composition, species richness and ground cover differed between plots. Principal component analysis (PCA) extracted ten soil attributes that were then correlated with vegetation data. The attributes extracted by PCA, in order of most common variance, were: water content, pH, terrolas depth, elevation, slope angle, leaf litter depth, total organic carbon, and counts of macrofauna, fungi and bacteria. All extracted attributes differed between plots, and all except bacteria correlated with at least one vegetation attribute. Water content and pH correlated most strongly with vegetation cover suggesting an increase in soil moisture and a reduction in pH are required in order to improve vegetation rehabilitation at Yarraman Mine. Further study is recommended to confirm these results using controlled experiments and to test potential solutions, such as organic amendments.

Mutans Streptococci and Lactobacilli Colonization in Predentate Children from the Neonatal Period to Seven Months of Age

Background/Aims: The aim of this study was to investigate the colonization of mutans streptococci (MS) and lactobacilli (LB) in predentate children from the neonatal period to 7 months. Methods: A total of 957 mother-and-child pairs were recruited from birth and followed up at 7 months. The 283 children who did not have erupted teeth at the second visit were included in the study. Oral mucosal swabs were taken, and the presence of MS and LB was determined using a commercial microbiological culture kit. Results: At mean ages of 34 days and 7 months, 9 and 11% of the infants, respectively, showed the presence of MS. In contrast, LB presence increased from 24 to 47% (p < 0.0001). MS presence in the neonatal period was associated with maternal MS counts of >105 CFU/ml (p = 0.05), while LB presence was associated with natural birth (p = 0.03) and maternal LB presence (p = 0.02). At 7 months, MS presence was associated with maternal MS counts (p = 0.02) and LB counts of >105 CFU/ml (p = 0.007). Additional predictors of MS presence at 7 months were a child’s MS counts of >105 CFU/ml at the neonatal visit (p = 0.019) and nighttime bottle feeding (p = 0.024). LB presence at 7 months was associated with maternal LB (p < 0.001) and MS presence (p = 0.02). Conclusions: MS and LB can be detected by culture in the oral cavity as early as 34 days after birth. Their infection rates increase to 11 and 47%, respectively, by the time the children reach the end of the predentate stage of oral development.

The relationship between Bayesian motor unit number estimation and histological measurements of motor neurons in wild-type and SOD1 G93A mice

Objective: To assess the relationship between Bayesian MUNE and histological motor neuron counts in wild-type mice and in an animal model of ALS. Methods: We performed Bayesian MUNE paired with histological counts of motor neurons in the lumbar spinal cord of wild-type mice and transgenic SOD1 G93A mice that show progressive weakness over time. We evaluated the number of acetylcholine endplates that were innervated by a presynaptic nerve. Results: In wild-type mice, the motor unit number in the gastrocnemius muscle estimated by Bayesian MUNE was approximately half the number of motor neurons in the region of the spinal cord that contains the cell bodies of the motor neurons supplying the hindlimb crural flexor muscles. In SOD1 G93A mice, motor neuron numbers declined over time. This was associated with motor endplate denervation at the end-stage of disease. Conclusion: The number of motor neurons in the spinal cord of wild-type mice is proportional to the number of motor units estimated by Bayesian MUNE. In SOD1 G93A mice, there is a lower number of estimated motor units compared to the number of spinal cord motor neurons at the end-stage of disease, and this is associated with disruption of the neuromuscular junction. Significance: Our finding that the Bayesian MUNE method gives estimates of motor unit numbers that are proportional to the numbers of motor neurons in the spinal cord supports the clinical use of Bayesian MUNE in monitoring motor unit loss in ALS patients. © 2012 International Federation of Clinical Neurophysiology.

Following fashion: Visible progressiveness and the social construction of firm value

A central tenet underlying studies on management fashions is that the diffusion of novel forms, models and techniques is driven by an institutional norm of progress, which is the societal expectation that managers will continuously use 'new and improved' management practices. We add to the literature on management fashions by arguing that, if the display of progressiveness in the manner of managing and organizing is expected of organizations, firms that are visibly progressive would be evaluated more positively by organizational audiences following this institutional prescription. Using article counts of co-occurrences of firms and various fashionable management practices in Wall Street Journal, we hypothesize positive effects of such associations on security analysts' evaluations of these firms. Results support this hypothesis. Our study enriches the management fashion literature by highlighting the consequential relevance of organizational adherence to the norm of progress.