Time series classification for the prediction of dialysis in critically ill patients using echo statenetworks

Objective: Time series often appear in medical databases, but only few machine learning methods exist that process this kind of data properly. Most modeling techniques have been designed with a static data model in mind and are not suitable for coping with the dynamic nature of time series. Recurrent neural networks (RNNs) are often used to process time series, but only a few training algorithms exist for RNNs which are complex and often yield poor results. Therefore, researchers often turn to traditional machine learning approaches, such as support vector machines (SVMs), which can easily be set up and trained and combine them with feature extraction (FE) and selection (FS) to process the high-dimensional temporal data. Recently, a new approach, called echo state networks (ESNs), has been developed to simplify the training process of RNNs. This approach allows modeling the dynamics of a system based on time series data in a straightforwardway.The objective of this study is to explore the advantages of using ESN instead of other traditional classifiers combined with FE and FS in classification problems in the intensive care unit (ICU) when the input data consists of time series. While ESNs have mostly been used to predict the future course of a time series, we use the ESN model for classification instead. Although time series often appear in medical data, little medical applications of ESNs have been studiedyet.Methods and material: ESN is used to predict the need for dialysis between the fifth and tenth day after admission in the ICU. The input time series consist of measured diuresis and creatinine values during the first 3days after admission. Data about 830 patients was used for the study, of which 82 needed dialysis between the fifth and tenth day after admission. ESN is compared to traditional classifiers, a sophisticated and a simple one, namely support vector machines and the naive Bayes (NB) classifier. Prior to the use of the SVM and NB classifier, FE and FS is required to reduce the number of input features and thus alleviate the curse dimensionality. Extensive feature extraction was applied to capture both the overall properties of the time series and the correlation between the different measurements in the time series. The feature selection method consists of a greedy hybrid filter-wrapper method using a NB classifier, which selects in each iteration the feature that improves prediction the best and shows little multicollinearity with the already selected set. Least squares regression with noise was used to train the linear readout function of the ESN to mitigate sensitivity to noise and overfitting. Fisher labeling was used to deal with the unbalanced data set. Parameter sweeps were performed to determine the optimal parameter values for the different classifiers. The area under the curve (AUC) and maximum balanced accuracy are used as performance measures. The required execution time was also measured.Results: The classification performance of the ESN shows significant difference at the 5% level compared to the performance of the SVM or the NB classifier combined with FE and FS. The NB+FE+FS, with an average AUC of 0.874, has the best classification performance. This classifier is followed by the ESN, which has an average AUC of 0.849. The SVM+FE+FS has the worst performance with an average AUC of 0.838. The computation time needed to pre-process the data and to train and test the classifier is significantly less for the ESN compared to the SVM andNB.Conclusion: It can be concluded that the use of ESN has an added value in predicting the need for dialysis through the analysis of time series data. The ESN requires significantly less processing time, needs no domain knowledge, is easy to implement, and can be configured using rules ofthumb.     

Estimation of the dynamics and yields of cereals in a semi-arid area using remote sensing and the SAFY growth model

In semi-arid areas, a strongly variable climate represents a major risk for food safety. An operational grain yield forecasting system, which could help decision-makers to make early assessments and plan annual imports, is thus needed. It can be challenging to monitor the crop canopy and production capacity of plants, especially cereals. In this context, the aim of the present study is to analyse the characteristics of two types of irrigated and non-irrigated cereals: barley and wheat. Through the use of a rich database, acquired over a period of two years for more than 30 test fields, and from 20 optical satellite SPOT/HRV images, two research approaches are considered. First, statistical analysis is used to characterize the vegetation’s dynamics and grain yield, based on remotely sensed (satellite) normalized difference vegetation index (NDVI) measurements. A relationship is established between the NDVI and LAI (leaf area index). Different robust relationships (exponential or linear) are established between the satellite NDVI index acquired from SPOT/HRV images, just before the time of maximum growth (April), and grain and straw, for barley and wheat vegetation covers. Following validation of the proposed empirical approaches, yield maps are produced for the studied site. The second approach is based on the application of a Simple Algorithm for Yield Estimation (SAFY) growth model, developed to simulate the dynamics of the LAI and the grain yield. An inter-comparison between ground yield measurements and SAFY model simulations reveals that yields are underestimated by this model. Finally, the combination of multi-temporal satellite measurements with the SAFY model estimations is also proposed for the purposes of yield mapping. Although the results produced by the SAFY model are found to be reasonably well correlated with those determined by satellite measurements (NDVI), the grain yields are nevertheless underestimated.     

Robust output feedback stabilization of uncertain time-varying state-delayed systems with saturating actuators

The robust output feedback stabilization problem for state-delayed systems with time-varying delays and saturating actuators is addressed here. The systems considered are continuous-time, with parametric uncertainties entering all the matrices in the system representation. A saturating control law is designed and a region of initial conditions is specified within which local asymptotic stability of the closed-loop system is ensured. The least conservative approach that employs the Lyapunov–Krasovskii functional is adopted to ensure stabilization. The designed controller is dependent on the time-delay and its rate of change. The controller is constructed in terms of the solution to a set of matrix inequalities.     

WIRELESS BODY-AREA NETWORKS

The last decade has witnessed the convergence of three giant worlds:electronics,computer science and telecommunications.The next decade should follow this convergence in most of our activities with the generalization of sensor networks.In particular with the progress in medicine,people live longer and the aging of population will push the development of wireless person-     

Encoding and selective activation of "metabolic memories" in the rat.

Experiment 1 used Pavlovian conditioning procedures to show that rats formed distinct memorial representations of 2 (peanut oil and sucrose pellets) unconditioned stimuli (USs) that could be activated by 2 different conditioned stimuli (CSs). After training in Experiment 2, rats injected with the lipid antimetabolite Na-2-mercaptoacetate (MA) responded more to the CS for oil than to the CS for sucrose. This pattern was not shown by rats that received isotonic saline or systemic 2-deoxy-d-glucose (a glucose antimetabolite). By contrast, intracerebroventricular infusion of the glucose antimetabolite 5-thioglucose selectively promoted responding to the CS for sucrose (Experiment 4). Thus, lipoprivic and glucoprivic treatments selectively promoted the activation of the memories of fat and carbohydrate USs, respectively. In Experiment 3, the capacity of MA to augment responding to a CS for oil was abolished for rats that received subdiaphragmatic vagal deafferentation. This indicates that the capacity of lipoprivic signals to selectively activate the representations of fat USs may depend on vagal afferent fibers. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Static Strategies for Worksharing with Unrecoverable Interruptions

One has a large workload that is “divisible”—its constituent work’s granularity can be adjusted arbitrarily—and one has access to p remote worker computers that can assist in computing the workload. How can one best utilize the workers? Complicating this question is the fact that each worker is subject to interruptions (of known likelihood) that kill all work in progress on it. One wishes to orchestrate sharing the workload with the workers in a way that maximizes the expected amount of work completed. Strategies are presented for achieving this goal, by balancing the desire to checkpoint often—thereby decreasing the amount of vulnerable work at any point—vs. the desire to avoid the context-switching required to checkpoint. Schedules must also temper the desire to replicate work, because such replication diminishes the effective remote workforce. The current study demonstrates the accessibility of strategies that provably maximize the expected amount of work when there is only one worker (the case p=1) and, at least in an asymptotic sense, when there are two workers (the case p=2); but the study strongly suggests the intractability of exact maximization for p≥2 computers, as work replication on multiple workers joins checkpointing as a vehicle for decreasing the impact of work-killing interruptions. We respond to that challenge by developing efficient heuristics that employ both checkpointing and work replication as mechanisms for decreasing the impact of work-killing interruptions. The quality of these heuristics, in expected amount of work completed, is assessed through exhaustive simulations that use both idealized models and actual trace data.     

Toxic effects of hexavalent chromium on brook trout (Salvelinus fontinalis) and rainbow trout (Salmo gairdneri)

Exposing brook trout to various concentrations of chromium [Cr(VI)] for up to 22 months (including reproduction) significantly increased alevin mortality at 0.35 mg Cr l⁻¹ and retarded growth of young brook trout at the lowest concentration tested (0.01 mg Cr l⁻¹). Eight month exposures of rainbow trout significantly increased alevin mortality at 0.34 mg Cr l⁻¹ and also retarded growth at the lowest concentration tested (0.10 mg Cr l⁻¹). Exposures of brook trout lasting 22 months showed, however, that growth was only temporarily affected, and therefore, it was not used as an end point to measure the affects of chromium on either species. Reproduction, and embryo hatchability of brook trout were unaffected at Cr(VI) concentrations that affected survival of newly hatched alevins. The maximum acceptable toxicant concentration (MATC) for brook and rainbow trout exposed to Cr(VI) in water with a hardness of 45 mg l⁻¹ (as CaCO₃) and a pH range of 7–8 lies between 0.20 and 0.35 mg Cr l⁻¹. The 96-h lc₅₀ for brook and rainbow trout was 59 and 69 mg Cr l⁻¹, respectively: therefore, the application factor (MATC/96-h lc₅₀) for both species lies between 0.003 and 0.006.     

Meeting the challenge of water sustainability: The role of process systems engineering

This white paper is the result of discussions during the FIPSE-4 conference ( http://fi-in-pse.org ) in June 2018. It aims to highlight open problems and provide directions for future research in the area of water with emphasis on its agricultural usages. Some of the open problems discussed are: (a) the use of ecosystems as unit operations to understand their role in providing freshwater and in cleaning polluted water; (b) consideration of interactions and independencies between flows of water and other resources, such as food, energy, materials, ecosystem services, and environmental emissions; (c) challenges in modeling, sensing, and closed-loop control in precision irrigation. In particular, the development of agro-hydrological models that balance computing speed versus solution details and accuracy: (d) The use of state and parameter estimation approaches, through field measurements, to obtain soil moisture levels accurately; and (e) decision support systems to administer water and nutrient needs for optimum yields of agricultural products.     

Effect of Wood Welding Process on Chemical Constituents of Australian Eucalyptus

Thermochemical changes occurring during wood welding were investigated in Eucalyptus saligna and Eucalyptus pilularis. Unwelded reference wood and material from welded interface were compared via Py-GC/MS, thermogravimetric analyses, X-ray photoelectron spectroscopy, and attenuated total reflection Fourier transform infrared spectroscopy to explain differences in mechanical properties of welded wood between species. It appeared that the species originally containing more condensed substructures also provided stronger joints. The condensation index after welding allowed validating that the adhesive properties of lignin are more accessible in such species. The presence of more carbonyl functions, attributed to carbohydrate solicitation during welding process, potentially made lignin less accessible. Changes following welding were evidenced by the contribution of extractive compounds, fatty acid chains, and terpenoids, possibly reacting with hydroxyl groups and leading to adhesive properties by chemical linkage through new covalent bonds formation. Results corroborate the better mechanical properties of E. saligna, providing stronger joints possibly due to more accessible adhesive properties by esterification between lignin and fatty acid.