Type Of Response Function Research Articles

A retrospective study of deep learning generalization across two centers and multiple models of X-ray devices using COVID-19 chest-X rays

Generalization of deep learning (DL) algorithms is critical for the secure implementation of computer-aided diagnosis systems in clinical practice. However, broad generalization remains to be a challenge in machine learning. This research aims to identify and study potential factors that can affect the internal validation and generalization of DL networks, namely the institution where the images come from, the image processing applied by the X-ray device, and the type of response function of the X-ray device. For these purposes, a pre-trained convolutional neural network (CNN) (VGG16) was trained three times for classifying COVID-19 and control chest radiographs with the same hyperparameters, but using different combinations of data acquired in two institutions by three different X-ray device manufacturers. Regarding internal validation, the addition of images from an external institution to the training set did not modify the algorithm’s internal performance, however, the inclusion of images acquired by a device from a different manufacturer decreased the performance up to 8% (p < 0.05). In contrast, generalization across institutions and X-ray devices with the same type of response function was achieved. Nonetheless, generalization was not observed across devices with different types of response function. This factor was the key impediment to achieving broad generalization in our research, followed by the device’s image-processing and the inter-institutional differences, which both reduced generalization performance to 18.9% (p < 0.05), and 9.8% (p < 0.05), respectively. Finally, clustering analysis with features extracted by the CNN was performed, revealing a substantial dependence of feature values extracted by the pre-trained CNN on the X-ray device which acquired the images.

Hutchinson model of Holling type II with delay in prey

The problems in daily life can be solved with mathematical models, such as the model of Lotka-Volterra to find out how growth of animals. This model can be modified into the Predator-Prey Model of Hutchinson with delay time and Response Function of Holling type II in which one as active predator and the other as prey, delay time is considered on prey population with assuming that supporting resources of prey are adequate and this type of response function is as interactions between populations of predator and prey. In this paper, linearization and eigenvalues are used to analyze the stability solution of the fixed point so that it can be known the effect of delay time and response function of this type. There are three fixed points are obtained without delay time with two fixed points are stable under certain conditions, but delay time can change the stability of the fixed point.

Do placebo based validation standards mimic real batch products behaviour? Case studies

Analytical methods validation is a mandatory step to evaluate the ability of developed methods to provide accurate results for their routine application. Validation usually involves validation standards or quality control samples that are prepared in placebo or reconstituted matrix made of a mixture of all the ingredients composing the drug product except the active substance or the analyte under investigation. However, one of the main concerns that can be made with this approach is that it may lack an important source of variability that come from the manufacturing process. The question that remains at the end of the validation step is about the transferability of the quantitative performance from validation standards to real authentic drug product samples. In this work, this topic is investigated through three case studies. Three analytical methods were validated using the commonly spiked placebo validation standards at several concentration levels as well as using samples coming from authentic batch samples (tablets and syrups). The results showed that, depending on the type of response function used as calibration curve, there were various degrees of differences in the results accuracy obtained with the two types of samples. Nonetheless the use of spiked placebo validation standards was showed to mimic relatively well the quantitative behaviour of the analytical methods with authentic batch samples. Adding these authentic batch samples into the validation design may help the analyst to select and confirm the most fit for purpose calibration curve and thus increase the accuracy and reliability of the results generated by the method in routine application.

Interpretation of environmental tracers in groundwater systems with stagnant water zones

Lumped-parameter models are commonly applied for determining the age of water from time records of transient environmental tracers. The simplest models (e.g. piston flow or exponential) are also applicable for dating based on the decay or accumulation of tracers in groundwater systems. The models are based on the assumption that the transit time distribution function (exit age distribution function) of the tracer particles in the investigated system adequately represents the distribution of flow lines and is described by a simple function. A chosen or fitted function (called the response function) describes the transit time distribution of a tracer which would be observed at the output (discharge area, spring, stream, or pumping wells) in the case of an instantaneous injection at the entrance (recharge area). Due to large space and time scales, response functions are not measurable in groundwater systems, therefore, functions known from other fields of science, mainly from chemical engineering, are usually used. The type of response function and the values of its parameters define the lumped-parameter model of a system. The main parameter is the mean transit time of tracer through the system, which under favourable conditions may represent the mean age of mobile water. The parameters of the model are found by fitting calculated concentrations to the experimental records of concentrations measured at the outlet. The mean transit time of tracer (often called the tracer age), whether equal to the mean age of water or not, serves in adequate combinations with other data for determining other useful parameters, e.g. the recharge rate or the content of water in the system. The transit time distribution and its mean value serve for confirmation or determination of the conceptual model of the system and/or estimation of its potential vulnerability to anthropogenic pollution. In the interpretation of environmental tracer data with the aid of the lumped-parameter models, the influence of diffusion exchange between mobile water and stagnant or quasi-stagnant water is seldom considered, though it leads to large differences between tracer and water ages. Therefore, the article is focused on the transit time distribution functions of the most common lumped-parameter models, particularly those applicable for the interpretation of environmental tracer data in double-porosity aquifers, or aquifers in which aquitard diffusion may play an important role. A case study is recalled for a confined aquifer in which the diffusion exchange with aquitard most probably strongly influenced the transport of environmental tracers. Another case study presented is related to the interpretation of environmental tracer data obtained from lysimeters installed in the unsaturated zone with a fraction of stagnant water.

Quality compromises incorporated in simplex optimisation of a flow injection system

A general method for experimental optimisation applied to the flow injection technique is presented. This method is based on the development of a particular type of response function through the area of a polygonal figure that it depends on four main variables: sensitivity, precision, cost and sample throughput. The optimisation was performed by processing these variables in a simultaneous way, following the principles of the Simplex method. Thus, the main parameters that define the overall analytical quality have been integrated, in order to give information about the quality compromises which characterise the quality policy of a particular laboratory. An evaluation of the advantages of this proposed method in comparison to the univariate optimisation methods has been also included, by using the method for the determination of furfural with aniline.

Toroidal order in crystals

Abstract A new ordered state characterized by non-zero toroidal moment density is described. A review of its electrical and magnetic properties is presented. A second order phase transition to the toroidal state is accompanied by Curie-Weiss behavior of a new type of response function, the toroidal susceptibility, which describes the response to an external current. The toroidal state can be either of spin or of orbital nature. Orbital toroidal magnetics can possess large diamagnetism.

Response function of the second kind in many-body systems

We analyze a new type of response function which portrays the properties of a system perturbed by an external field in terms of the perturbed two-point correlations of density fluctuations rather than in terms of perturbed averages of physical quantities. This “response function of the second kind” satisfies both fluctuation-dissipation-like theorems, relating it to three-point equilibrium functions, and hierarchical relationship linking it to conventional quadratic (rather than linear) response functions. In the equal-time limit, when the two density fluctuations are observed at the same time, the response function of the second kind is intimately connected to the two-particle correlation function of kinetic theory. This linkage opens an avenue for developing novel approximation techniques for correlated many-body systems.