Multi-scale support vector algorithms for hot spot detection and modelling

The algorithmic approach to data modelling has developed rapidly these last years, in particular methods based on data mining and machine learning have been used in a growing number of applications. These methods follow a data-driven methodology, aiming at providing the best possible generalization and predictive abilities instead of concentrating on the properties of the data model. One of the most successful groups of such methods is known as Support Vector algorithms. Following the fruitful developments in applying Support Vector algorithms to spatial data, this paper introduces a new extension of the traditional support vector regression (SVR) algorithm. This extension allows for the simultaneous modelling of environmental data at several spatial scales. The joint influence of environmental processes presenting different patterns at different scales is here learned automatically from data, providing the optimum mixture of short and large-scale models. The method is adaptive to the spatial scale of the data. With this advantage, it can provide efficient means to model local anomalies that may typically arise in situations at an early phase of an environmental emergency. However, the proposed approach still requires some prior knowledge on the possible existence of such short-scale patterns. This is a possible limitation of the method for its implementation in early warning systems. The purpose of this paper is to present the multi-scale SVR model and to illustrate its use with an application to the mapping of Cs137 activity given the measurements taken in the region of Briansk following the Chernobyl accident.

Ultra high performance supercritical fluid chromatography coupled with tandem mass spectrometry for screening of doping agents. II: Analysis of biological samples.

The potential and applicability of UHPSFC-MS/MS for anti-doping screening in urine samples were tested for the first time. For this purpose, a group of 110 doping agents with diverse physicochemical properties was analyzed using two separation techniques, namely UHPLC-MS/MS and UHPSFC-MS/MS in both ESI+ and ESI- modes. The two approaches were compared in terms of selectivity, sensitivity, linearity and matrix effects. As expected, very diverse retentions and selectivities were obtained in UHPLC and UHPSFC, proving a good complementarity of these analytical strategies. In both conditions, acceptable peak shapes and MS detection capabilities were obtained within 7min analysis time, enabling the application of these two methods for screening purposes. Method sensitivity was found comparable for 46% of tested compounds, while higher sensitivity was observed for 21% of tested compounds in UHPLC-MS/MS and for 32% in UHPSFC-MS/MS. The latter demonstrated a lower susceptibility to matrix effects, which were mostly observed as signal suppression. In the case of UHPLC-MS/MS, more serious matrix effects were observed, leading typically to signal enhancement and the matrix effect was also concentration dependent, i.e., more significant matrix effects occurred at the lowest concentrations.

Rapid high-performance liquid chromatographic determination with fluorescence detection of furosemide in human body fluids and its confirmation by gas chromatography-mass spectrometry.

Furosemide (FD: Lasix) is a loop diuretic which strongly increases both urine flow and electrolyte urinary excretion. Healthy volunteers were administered 40 mg orally (dissolved in water) and concentrations of FD were determined in serum and urine for up to 6 h for eight subjects, who absorbed water at a rate of 400 ml/h. Quantification was performed by HPLC with fluorescence detection (excitation at 233 nm, emission at 389 nm) with a limit of detection of 5 ng/ml for a 300-microliters sample. The elution of FD was completed within 4 min using a gradient of acetonitrile concentration rising from 30 to 50% in 0.08 M phosphoric acid. The delay to the peak serum concentration ranged from 60 to 120 min. FD was still easily measurable in the sera from all subjects 6 h after administration. In urine, the excretion rates reached their maximum between 1 and 3 h. The total amount of FD excreted in the urine averaged 11.2 mg (range 7.6-14.0 mg), with a mean urine volume of 3024 ml (range 2620-3596 ml). Moreover, the urine density was lower than 1.010 (recommended as an upper limit in doping analysis to screen diuretics) only for 2 h. An additional volunteer was administered 40 mg of FD and his urine was collected over a longer period. FD was still detectable 48 h after intake. Gas chromatography-mass spectrometry with different types of ionization was used to confirm the occurrence of FD after permethylation of the extract. Negative-ion chemical ionization, with ammonia as reactant gas, was found to be the most sensitive method of detection.