The Effect of Spatial Scale on Predicting Time Series: A Study on Epidemiological System Identification

A susceptible-infective-recovered (SIR) epidemiological model based on probabilistic cellular automaton (PCA) is employed for simulating the temporal evolution of the registered cases of chickenpox in Arizona, USA, between 1994 and 2004. At each time step, every individual is in one of the states S, I, or R. The parameters of this model are the probabilities of each individual (each cell forming the PCA lattice ) passing from a state to another state. Here, the values of these probabilities are identified by using a genetic algorithm. If nonrealistic values are allowed to the parameters, the predictions present better agreement with the historical series than if they are forced to present realistic values. A discussion about how the size of the PCA lattice affects the quality of the model predictions is presented. Copyright (C) 2009 L. H. A. Monteiro et al.

Ramp and constant power trials produce equivalent critical power estimates

The standard critical power test protocol on the cycle prescribes a series of trials to exhaustion, each at a different but constant power setting. Recently the protocol has been modified and applied to a series of trials to exhaustion each at a different ramp incremental rate. This study was undertaken to compare critical power and anaerobic work capacity estimates in the same group of subjects when derived from the two protocols. Ten male subjects of mixed athletic ability cycled to exhaustion on eight occasions in randomized order over a 3-wk period. Four trials were performed at differing constant power settings and four trials on differing ramp incremental rates. Both critical power and anaerobic work capacity were estimated for each subject by curve fitting of the ramp model and of three versions of the constant power model. After adjusting for inter-subject variability, no significant differences were detected between critical power estimates or between anaerobic work capacity estimates from any model formulation or from the two protocols. It is concluded that both the ramp and constant power protocols produce equivalent estimates for critical power and anaerobic work capacity.

Influence of hysteresis on tidal capillary fringe dynamics in a well-sorted sand

Nielsen and Perrochet [Adv. Water Resour. 23 (2000) 503] presented experimental data for cyclic water movement in the vadose zone above an oscillating watertable. The response of the watertable to cyclic forcing was characterised by the ratios of the forcing head to watertable amplitudes and their associated phase lag. They found that their non-hysteretic Richards' equation model failed to represent the observed behaviour of these parameters. This paper explores the effect on the simulated capillary fringe dynamics (in terms of these parameters) of including varying degrees of hysteresis in the moisture retention curve used in a numerical model of their experiment. It is clear that hysteresis can indeed account for observed discrepancies between simulation and experiment and that the effect of hysteresis varies with the frequency of oscillation. The use of a single-valued mean retention curve, as advocated by some authors, fails to provide a match between the simulated and observed behaviour of the Nielsen and Perrochet parameters, but is shown to be adequate for predicting time-averaged soil moisture profiles. (C) 2003 Elsevier Ltd. All rights reserved.

Infections and functional impairment in nursing home residents: a reciprocal relationship.

OBJECTIVES: To determine the relationship between infections and functional impairment in nursing home residents. DESIGN: Prospective cohort study (follow-up period, 6 months). SETTING: Thirty-nine nursing homes in western Switzerland. PARTICIPANTS: A total of 1,324 residents aged 65 and older (mean age 85.7; 76.6% female) who agreed to participate, or their proxies, by oral informed consent. MEASUREMENTS: Functional status measured every 3 months. Two different outcomes were used: (a) functional decline defined as death or decreased function at follow-up and (b) functional status score using a standardized measure. RESULTS: At the end of follow-up, mortality was 14.6%, not different for those with and without infection (16.2% vs 13.1%, P=.11). During both 3-month periods, subjects with infection had higher odds of functional decline, even after adjustment for baseline characteristics and occurrence of a new illness (adjusted odds ratio (AOR)=1.6, 95% confidence interval (CI)=1.2-2.2, P=.002, and AOR=1.5, 95% CI=1.1-2.0, P=.008, respectively). The odds of decline increased in a stepwise fashion in patients with zero, one, and two or more infections. The analyses predicting functional status score (restricted to subjects who survived) gave similar results. A survival analysis predicting time to first infection confirmed a stepwise greater likelihood of infection in subjects with moderate and severe impairment at baseline than in subjects with no or mild functional impairment at baseline. CONCLUSION: Infections appear to be both a cause and a consequence of functional impairment in nursing home residents. Further studies should be undertaken to investigate whether effective infection control programs can also contribute to preventing functional decline, an important component of these residents' quality of life.

Infections and functionnal impairment in nursing home residents : a reciprocal relationship

RESUME: Contexte : l'objectif de cette étude de cohorte prospective était de déterminer la relation entre la survenue d'infections et la dépendance fonctionnelle chez des résidents d'établissements de long séjour durant une période de 6 mois. Population et méthode : les patients inclus (1324 résidents) étaient âgés de 65 ans et plus (âge moyen 85.7 ans, 76.6% de femmes), étaient des résidents de 39 EMS du canton de Vaud. Au baseline, des données démographiques, médicales, concernant les facteurs de risque et protecteurs des infections ont été récoltées. Au cours du suivi de 6 mois, les infirmières des EMS ont documenté la survenue de symptômes et signes d'infection en utilisant les critères développés spécifiquement par l'APIC pour les établissements de long séjour. Les mesures du status fonctionnel ont été évaluées au baseline, à 3 mois et à 6 mois. Deux outcomes différents ont été utilisés : a) le déclin fonctionnel défini comme le décès ou une diminution des capacités fonctionnelles au suivi, b) le status fonctionnel mesuré par une échelle standardisée. Résultats : à la fin du suivi, la mortalité était de 14.6%, similaire pour les résidents avec et sans infection (16.2% versus 13.1%, P .11). Durant les 2 périodes de suivi de 3 mois, les sujets ayant présenté une ou plusieurs infections avaient des odds de déclin fonctionnel plus élevés, y compris après ajustement pour les caractéristiques démographiques, médicales et fonctionnelles du baseline, ainsi que la survenue de nouvelles maladies (odds ratio ajustés (OR) = 1.6, intervalle de confiance à 95% (IC) = 1.2-2.2, P = .002 et OR = 1.5, 95% IC= 1.1-2.0, P= .008, respectivement). Comparés aux résidents non infectés, les odds de déclin fonctionnel augmentaient significativement et graduellement chez ceux ayant eu une, respectivement 2 infections ou plus. L'analyse prédisant le score fonctionnel (restreinte aux sujets ayant survécu) a donné des résultats similaires. Finalement, une analyse de survie prédisant le temps jusqu'à la première infection a confirmé une augmentation progressive de la probabilité d'infection chez les sujets avec dépendance fonctionnelle modérée, respectivement sévère, comparés aux sujets indépendants à la ligne de base. Conclusion : chez les résidents de long séjour, les infections sont à la fois cause et conséquence de la dépendance fonctionnelle. Des études futures devraient être entreprises pour investiguer si des programmes de prévention des infections peuvent également contribuer à prévenir le déclin fonctionnel, un facteur important pour la qualité de vie de ces résidents. ABSTRACT: Objectives: To determine the relationship between infections and functional impairment in nursing home residents. Design: Prospective cohort study (follow-up period, 6 months). Setting: Thirty-nine nursing homes in western Switzerland. Participants: A total of 1,324 residents aged 65 and older (mean age 85.7; 76.6% female) who agreed to participate, or their proxies, by oral informed consent. Measurements: Functional status measured every 3 months. Two different outcomes were used: (a) functional decline defined as death or decreased function at follow-up and (b) functional status score using a standardized measure. Results: At the end of follow-up, mortality was 14.6%, not different for those with and without infection (16.2% vs 13.1%, P= .11) During both 3-month periods, subjects with infection had higher odds of functional decline, even after adjustment for baseline characteristics and occurrence of a new illness (adjusted odds ratio (AOR) = 1.6, 95% confidence interval (CI) = 1.2-2.2, P = .002, and AOR 1.5, 95% CI 1.1-2.0, P .008, respectively). The odds of decline increased in a stepwise fashion in patients with zero, one, and two or more infections. The analyses predicting functional status score (restricted to subjects who survived) gave similar results. A survival analysis predicting time to first infection confirmed a stepwise greater likelihood of infection in subjects -with moderate and severe impairment at baseline than in subjects with no or mild functional impairment at baseline. Conclusion: Infections appear to be both a cause and a consequence of functional impairment in nursing home residents. Further studies should be undertaken to investigate whether effective infection control programs can also contribute to preventing functional decline, an important component of these residents' quality of life.

A universal agro-hydrological model for water and nitrogen cycles in the soil-crop system SMCR_N: critical update and further validation

Agro-hydrological models have widely been used for optimizing resources use and minimizing environmental consequences in agriculture. SMCRN is a recently developed sophisticated model which simulates crop response to nitrogen fertilizer for a wide range of crops, and the associated leaching of nitrate from arable soils. In this paper, we describe the improvements of this model by replacing the existing approximate hydrological cascade algorithm with a new simple and explicit algorithm for the basic soil water flow equation, which not only enhanced the model performance in hydrological simulation, but also was essential to extend the model application to the situations where the capillary flow is important. As a result, the updated SMCRN model could be used for more accurate study of water dynamics in the soil-crop system. The success of the model update was demonstrated by the simulated results that the updated model consistently out-performed the original model in drainage simulations and in predicting time course soil water content in different layers in the soil-wheat system. Tests of the updated SMCRN model against data from 4 field crop experiments showed that crop nitrogen offtakes and soil mineral nitrogen in the top 90 cm were in a good agreement with the measured values, indicating that the model could make more reliable predictions of nitrogen fate in the crop-soil system, and thus provides a useful platform to assess the impacts of nitrogen fertilizer on crop yield and nitrogen leaching from different production systems. (C) 2010 Elsevier B.V. All rights reserved.

SURVIVAL PREDICTION FOR BRAIN TUMOR PATIENTS USING GENE EXPRESSION DATA

Brain tumor is one of the most aggressive types of cancer in humans, with an estimated median survival time of 12 months and only 4% of the patients surviving more than 5 years after disease diagnosis. Until recently, brain tumor prognosis has been based only on clinical information such as tumor grade and patient age, but there are reports indicating that molecular profiling of gliomas can reveal subgroups of patients with distinct survival rates. We hypothesize that coupling molecular profiling of brain tumors with clinical information might improve predictions of patient survival time and, consequently, better guide future treatment decisions. In order to evaluate this hypothesis, the general goal of this research is to build models for survival prediction of glioma patients using DNA molecular profiles (U133 Affymetrix gene expression microarrays) along with clinical information. First, a predictive Random Forest model is built for binary outcomes (i.e. short vs. long-term survival) and a small subset of genes whose expression values can be used to predict survival time is selected. Following, a new statistical methodology is developed for predicting time-to-death outcomes using Bayesian ensemble trees. Due to a large heterogeneity observed within prognostic classes obtained by the Random Forest model, prediction can be improved by relating time-to-death with gene expression profile directly. We propose a Bayesian ensemble model for survival prediction which is appropriate for high-dimensional data such as gene expression data. Our approach is based on the ensemble "sum-of-trees" model which is flexible to incorporate additive and interaction effects between genes. We specify a fully Bayesian hierarchical approach and illustrate our methodology for the CPH, Weibull, and AFT survival models. We overcome the lack of conjugacy using a latent variable formulation to model the covariate effects which decreases computation time for model fitting. Also, our proposed models provides a model-free way to select important predictive prognostic markers based on controlling false discovery rates. We compare the performance of our methods with baseline reference survival methods and apply our methodology to an unpublished data set of brain tumor survival times and gene expression data, selecting genes potentially related to the development of the disease under study. A closing discussion compares results obtained by Random Forest and Bayesian ensemble methods under the biological/clinical perspectives and highlights the statistical advantages and disadvantages of the new methodology in the context of DNA microarray data analysis.

Predicting triage waiting time in maternity emergency care by means of data mining

Healthcare organizations often benefit from information technologies as well as embedded decision support systems, which improve the quality of services and help preventing complications and adverse events. In Centro Materno Infantil do Norte (CMIN), the maternal and perinatal care unit of Centro Hospitalar of Oporto (CHP), an intelligent pre-triage system is implemented, aiming to prioritize patients in need of gynaecology and obstetrics care in two classes: urgent and consultation. The system is designed to evade emergency problems such as incorrect triage outcomes and extensive triage waiting times. The current study intends to improve the triage system, and therefore, optimize the patient workflow through the emergency room, by predicting the triage waiting time comprised between the patient triage and their medical admission. For this purpose, data mining (DM) techniques are induced in selected information provided by the information technologies implemented in CMIN. The DM models achieved accuracy values of approximately 94% with a five range target distribution, which not only allow obtaining confident prediction models, but also identify the variables that stand as direct inducers to the triage waiting times.

Predicting pre-triage waiting time in a maternity emergency room through data mining

An unsuitable patient flow as well as prolonged waiting lists in the emergency room of a maternity unit, regarding gynecology and obstetrics care, can affect the mother and child’s health, leading to adverse events and consequences regarding their safety and satisfaction. Predicting the patients’ waiting time in the emergency room is a means to avoid this problem. This study aims to predict the pre-triage waiting time in the emergency care of gynecology and obstetrics of Centro Materno Infantil do Norte (CMIN), the maternal and perinatal care unit of Centro Hospitalar of Oporto, situated in the north of Portugal. Data mining techniques were induced using information collected from the information systems and technologies available in CMIN. The models developed presented good results reaching accuracy and specificity values of approximately 74% and 94%, respectively. Additionally, the number of patients and triage professionals working in the emergency room, as well as some temporal variables were identified as direct enhancers to the pre-triage waiting time. The imp lementation of the attained knowledge in the decision support system and business intelligence platform, deployed in CMIN, leads to the optimization of the patient flow through the emergency room and improving the quality of services.

Predicting soil erosion using Rusle and NDVI time series from TM Landsat 5

The objective of this work was to evaluate the seasonal variation of soil cover and rainfall erosivity, and their influences on the revised universal soil loss equation (Rusle), in order to estimate watershed soil losses in a temporal scale. Twenty-two TM Landsat 5 images from 1986 to 2009 were used to estimate soil use and management factor (C factor). A corresponding rainfall erosivity factor (R factor) was considered for each image, and the other factors were obtained using the standard Rusle method. Estimated soil losses were grouped into classes and ranged from 0.13 Mg ha-1 on May 24, 2009 (dry season) to 62.0 Mg ha-1 on March 11, 2007 (rainy season). In these dates, maximum losses in the watershed were 2.2 and 781.5 Mg ha-1 , respectively. Mean annual soil loss in the watershed was 109.5 Mg ha-1 , but the central area, with a loss of nearly 300.0 Mg ha-1 , was characterized as a site of high water-erosion risk. The use of C factor obtained from remote sensing data, associated to corresponding R factor, was fundamental to evaluate the soil erosion estimated by the Rusle in different seasons, unlike of other studies which keep these factors constant throughout time.

Predicting beef carcass retail products of Mediterranean buffaloes by real-time ultrasound measures

Twenty eight Mediterranean buffaloes bulls were scanned with real-time ultrasound (RTU), slaughtered, and fabricated into retail cuts to determine the potential for ultrasound measures to predict carcass retail yield. Ultrasound measures of fat thickness, ribeye area and rump fat thickness were recorded three to five days prior to slaughter. Carcass measurements were taken, and one side of each carcass was fabricated into retail cuts. Stepwise regression analysis was used to compare possible models for prediction of either kilograms or percent retail product from carcass mesaurements and ultrasound measures. Results indicate that possible prediction models for percent or kilograms of retail products using RTU measures were similar in their predictive power and accuracy when compared to models derived from carcass measurements. Both fat thickness and ribeye area were over-predicted when measured ultrasonically compared to measurements taken on the carcass in the cooler. The mean absolute differences for both traits are larger than the mean differences, indicating that some images were interpreted to be larger and some smaller than actual carcass measurements. Ultrasound measurements of REA and FT had positive correlations with carcass measures of the same traits (r=.96 for REA and r=.99 for FT). Standard errors of prediction currently are being used as the standard to certify ultrasound technicians for accuracy. Regression equations using live weight (LW), rib eye area (REAU) and subcutaneous fat thickness (FTU) between 12(th) and 13 (th) ribs and also over the biceps femoris muscle (FTP8) by ultrasound explained 95% of the variation in the hot carcass weight when measure immediately before slaughter.

Change over time in posttraumatic stress caused by myocardial infarction and predicting variables

The traumatic experience of a heart attack may evolve into symptoms of posttraumatic stress disorder, which can be diagnosed at the earliest 1 month after myocardial infarction (MI). While several predictors of posttraumatic stress in the first year after MI have been described, we particularly sought to identify longer-term predictors and predictors of change in posttraumatic stress over time.

Quantifying and Comparing the Accuracy of Binary Biomarkers When Predicting a Failure Time Outcome

The positive and negative predictive value are standard measures used to quantify the predictive accuracy of binary biomarkers when the outcome being predicted is also binary. When the biomarkers are instead being used to predict a failure time outcome, there is no standard way of quantifying predictive accuracy. We propose a natural extension of the traditional predictive values to accommodate censored survival data. We discuss not only quantifying predictive accuracy using these extended predictive values, but also rigorously comparing the accuracy of two biomarkers in terms of their predictive values. Using a marginal regression framework, we describe how to estimate differences in predictive accuracy and how to test whether the observed difference is statistically significant.

TIME SERIES ANALYSIS AS INPUT FOR PREDICTIVE MODELING: PREDICTING CARDIAC ARREST IN A PEDIATRIC INTENSIVE CARE UNIT

The first manuscript, entitled "Time-Series Analysis as Input for Clinical Predictive Modeling: Modeling Cardiac Arrest in a Pediatric ICU" lays out the theoretical background for the project. There are several core concepts presented in this paper. First, traditional multivariate models (where each variable is represented by only one value) provide single point-in-time snapshots of patient status: they are incapable of characterizing deterioration. Since deterioration is consistently identified as a precursor to cardiac arrests, we maintain that the traditional multivariate paradigm is insufficient for predicting arrests. We identify time series analysis as a method capable of characterizing deterioration in an objective, mathematical fashion, and describe how to build a general foundation for predictive modeling using time series analysis results as latent variables. Building a solid foundation for any given modeling task involves addressing a number of issues during the design phase. These include selecting the proper candidate features on which to base the model, and selecting the most appropriate tool to measure them. We also identified several unique design issues that are introduced when time series data elements are added to the set of candidate features. One such issue is in defining the duration and resolution of time series elements required to sufficiently characterize the time series phenomena being considered as candidate features for the predictive model. Once the duration and resolution are established, there must also be explicit mathematical or statistical operations that produce the time series analysis result to be used as a latent candidate feature. In synthesizing the comprehensive framework for building a predictive model based on time series data elements, we identified at least four classes of data that can be used in the model design. The first two classes are shared with traditional multivariate models: multivariate data and clinical latent features. Multivariate data is represented by the standard one value per variable paradigm and is widely employed in a host of clinical models and tools. These are often represented by a number present in a given cell of a table. Clinical latent features derived, rather than directly measured, data elements that more accurately represent a particular clinical phenomenon than any of the directly measured data elements in isolation. The second two classes are unique to the time series data elements. The first of these is the raw data elements. These are represented by multiple values per variable, and constitute the measured observations that are typically available to end users when they review time series data. These are often represented as dots on a graph. The final class of data results from performing time series analysis. This class of data represents the fundamental concept on which our hypothesis is based. The specific statistical or mathematical operations are up to the modeler to determine, but we generally recommend that a variety of analyses be performed in order to maximize the likelihood that a representation of the time series data elements is produced that is able to distinguish between two or more classes of outcomes. The second manuscript, entitled "Building Clinical Prediction Models Using Time Series Data: Modeling Cardiac Arrest in a Pediatric ICU" provides a detailed description, start to finish, of the methods required to prepare the data, build, and validate a predictive model that uses the time series data elements determined in the first paper. One of the fundamental tenets of the second paper is that manual implementations of time series based models are unfeasible due to the relatively large number of data elements and the complexity of preprocessing that must occur before data can be presented to the model. Each of the seventeen steps is analyzed from the perspective of how it may be automated, when necessary. We identify the general objectives and available strategies of each of the steps, and we present our rationale for choosing a specific strategy for each step in the case of predicting cardiac arrest in a pediatric intensive care unit. Another issue brought to light by the second paper is that the individual steps required to use time series data for predictive modeling are more numerous and more complex than those used for modeling with traditional multivariate data. Even after complexities attributable to the design phase (addressed in our first paper) have been accounted for, the management and manipulation of the time series elements (the preprocessing steps in particular) are issues that are not present in a traditional multivariate modeling paradigm. In our methods, we present the issues that arise from the time series data elements: defining a reference time; imputing and reducing time series data in order to conform to a predefined structure that was specified during the design phase; and normalizing variable families rather than individual variable instances. The final manuscript, entitled: "Using Time-Series Analysis to Predict Cardiac Arrest in a Pediatric Intensive Care Unit" presents the results that were obtained by applying the theoretical construct and its associated methods (detailed in the first two papers) to the case of cardiac arrest prediction in a pediatric intensive care unit. Our results showed that utilizing the trend analysis from the time series data elements reduced the number of classification errors by 73%. The area under the Receiver Operating Characteristic curve increased from a baseline of 87% to 98% by including the trend analysis. In addition to the performance measures, we were also able to demonstrate that adding raw time series data elements without their associated trend analyses improved classification accuracy as compared to the baseline multivariate model, but diminished classification accuracy as compared to when just the trend analysis features were added (ie, without adding the raw time series data elements). We believe this phenomenon was largely attributable to overfitting, which is known to increase as the ratio of candidate features to class examples rises. Furthermore, although we employed several feature reduction strategies to counteract the overfitting problem, they failed to improve the performance beyond that which was achieved by exclusion of the raw time series elements. Finally, our data demonstrated that pulse oximetry and systolic blood pressure readings tend to start diminishing about 10-20 minutes before an arrest, whereas heart rates tend to diminish rapidly less than 5 minutes before an arrest.