Mixed Diffusion Process Research Articles

Reliable generation of privacy-preserving synthetic electronic health record time series via diffusion models.

Electronic health records (EHRs) are rich sources of patient-level data, offering valuable resources for medical data analysis. However, privacy concerns often restrict access to EHRs, hindering downstream analysis. Current EHR deidentification methods are flawed and can lead to potential privacy leakage. Additionally, existing publicly available EHR databases are limited, preventing the advancement of medical research using EHR. This study aims to overcome these challenges by generating realistic and privacy-preserving synthetic EHRs time series efficiently. We introduce a new method for generating diverse and realistic synthetic EHR time series data using denoizing diffusion probabilistic models. We conducted experiments on 6 databases: Medical Information Mart for Intensive Care III and IV, the eICU Collaborative Research Database (eICU), and non-EHR datasets on Stocks and Energy. We compared our proposed method with 8 existing methods. Our results demonstrate that our approach significantly outperforms all existing methods in terms of data fidelity while requiring less training effort. Additionally, data generated by our method yield a lower discriminative accuracy compared to other baseline methods, indicating the proposed method can generate data with less privacy risk. The proposed model utilizes a mixed diffusion process to generate realistic synthetic EHR samples that protect patient privacy. This method could be useful in tackling data availability issues in the field of healthcare by reducing barrier to EHR access and supporting research in machine learning for health. The proposed diffusion model-based method can reliably and efficiently generate synthetic EHR time series, which facilitates the downstream medical data analysis. Our numerical results show the superiority of the proposed method over all other existing methods.

Detailed study of Safranin-O adsorption on sepiolite clay: Kinetics, thermodynamics, isotherms and theoretical calculations for optimal water treatment efficiency

Large-scale pollution requires industrial waste management policy. This study describes the adsorptive removal process for the cationic dye Safranin-O (Sfr), known for its high toxicity, using a natural, chemically unmodified clay classified as a fibrous clay, namely sepiolite. This adsorbent was thoroughly characterized using a variety of techniques, including Fourier transform infrared spectroscopy (FTIR), scanning electron microscopy with X-ray energy dispersive spectroscopy (SEM-EDS), X-ray diffraction (XRD), X-ray fluorescence (FX), zero charge point analysis (pHZCP) and Zeta potential. The kinetic study showed that adsorption followed pseudo-second-order kinetics associated with a mixed diffusion process including bulk and intra-particle diffusion. The adsorption process reached equilibrium within 300 min. Sorption isotherm models fitted the experimental data in the order: M-J (four parameters) > F-S V (five parameters) > F-S IV (four parameters) > Sips (three parameters) ≈ Hill (three parameters) > R-P (three parameters) > Freundlich (2 parameters) > Langmuir (2 parameters). The maximum adsorption capacity estimated by the Langmuir model was approximately 233.81 mg/g. Analysis of the thermodynamic parameters highlighted adsorption as a spontaneous, physical and exothermic process by nature, underlining the preponderance of randomness at the adsorbent surface. In addition, DFT calculations followed by some electronic parameters linked to the Sfr dye were studied. Topological characteristics such as Localised Orbital Locator (LOL), Electron Localisation Function (ELF), Visualizing Molecular Dynamics (VMD), and Reduced Density Gradient (RDG) were discussed using the multiwave function (Multiwfn) in liquid phase. Monte Carlo (MC) simulation models were conducted to probe the interaction and mechanism between the fibrous clay fraction surface and the Sfr dye. Overall, sepiolite, due to its abundance, low cost, efficiency and environmental safety, is emerging as a promising adsorbent for the treatment of wastewater containing harmful dyes.

Removal of toxic methyl orange by a cost-free and eco-friendly adsorbent: Mechanism, phytotoxicity, thermodynamics, and kinetics

This article describes the adsorptive removal of acute toxic methyl orange (MO) dye using natural jackfruit leaves powder (JLP) adsorbent without any chemical modification. The adsorbent was characterized by Fourier transform infrared spectroscopy (FTIR), scanning electron microscope equipped with X-ray energy dispersive spectroscopy (SEM-EDS), X-ray diffraction (XRD), thermogravimetric analysis (TGA), and point of zero charge (pHZPC) Analyses. Under the optimized condition, the highest removal efficiency of MO dye was found to be 78.10%. The adsorption kinetic models ensued pseudo-second order kinetics (R2 > 0.99) and mixed diffusion process (bulk and intra-particle diffusion). The Langmuir model was well fitted (R2 > 0.99) compared to other isotherms and indicated exothermic monolayer adsorption. The study of thermodynamic parameters revealed that the adsorption was spontaneous, physisorption, and exothermic in nature. Phytotoxicity trials on pea (Pisum sativum) seeds showed 0%, 76%, 82%, and 100% germination in experimental MO dye, treated dye, dye-loaded JLP adsorbent, and freshwater respectively, confirmed the nontoxic nature of treated dye as well as dye-loaded JLP adsorbent. Overall, simple preparation, abundant, cost-free, efficient, and environmentally benign nature make JLP a promising adsorbent for the remediation of harmful MO and other dyes contained wastewater.

Oxidation behavior of Ti–46Al–8Nb alloy with boron and carbon addition under thermal cycling conditions

Oxidation behavior of Ti–46Al–8Nb (in at.%) alloy with boron and carbon addition under thermal cycling conditions was investigated. Oxidation of Ti–46Al–8Nb, Ti–46Al–8Nb–1B and Ti–46Al–8Nb–1B-0.25C (in at.%) alloys was carried out at 1073 K in laboratory air for 42 cycles (1 cycle, 24 h), 1008 h in total. The mass loss rates of Ti–46Al–8Nb and Ti–46Al–8Nb–1B measured during the oxidation were similar. The best oxidation resistance was found for Ti–46Al–8Nb–1B-0.25C alloy with the smallest mass change. XRD and SEM-EDS investigations showed that in all cases, the oxide scales compositions were substantially similar. The scale consisted of an outer layer built of Al2O3 with the presence of some amounts of TiO2, an intermediate layer of the scale consisting of TiO2, an inner layer composed of oxides and nitrides. Additionally, niobium rich particles at the scale/substrate interface were present. The oxidation mechanism of Ti–46Al–8Nb was studied via two-stage isothermal oxidation (24 h in 16O2 followed by 24 h in 18O2) at 1073 K combined with secondary neutral mass spectroscopy (SNMS). These results indicate that the oxidation mechanism depends on a mixed diffusion process, consisting of outward transport of cations and simultaneous oxygen inward transport.

Characteristics of the mass transfer of phosphatidylcholine during its sorption on mesoporous composites based on MCM-41

The kinetic parameters of sorption of phosphatidylcholine on mesoporous composites based on MCM-41 are considered. It is noted that the possibility of both the diffusion and adsorption rate limitations of the process should be taken into account in the description of the kinetics of sorption of non-polar fat-soluble physiologically active compounds (PACs) from hexane solutions onto mesoporous materials of MCM- 41 type. The adequacy of using the Boyd diffusion model and the Lagergren, Ho and McKay, and Elovich models to describe the kinetics of sorption of phosphatidylcholine on mesoporous composites based on MCM-41 is shown. The contributions from diffusion limitation (internal and external) and the rate of the chemical step of adsorption to the overall rate of the sorption process are determined. It is found that the sorption of the phospholipid is a mixed diffusion process.

Development of Equivalent Circuits for Litihum-Air Battery Impedance Using Mixed and Isolated Diffusion Models

Typically, equivalent circuit models of electrochemical impedance spectroscopy (EIS) measurements include components that describe kinetic and diffusion behavior. Often, they may be depicted as a series combination of a charge-transfer resistance and a diffusion impedance element to represent a mixed diffusion process. Alternatively, diffusion-independent schematics also model impedance of electrochemical systems. Because equivalent circuit modeling is an inexact description of the electrochemical impedance, either schematic may or may not indicate a good fit of the measured spectra, depending on the parameters of the circuit components. The figure below shows the significance of a mixed diffusion circuit model versus a diffusion-independent circuit model fitting the impedance of a twice cycled lithium-air battery. An appropriate circuit model must not only fit the EIS measurement, but also consider the design of the physical system to account for possible side reactions that introduce ambiguity to the model. To overcome the impact of model uncertainty, however, alternative analysis techniques can help to verify the model of choice, such as the graphical analysis [1]. A graphical analysis of the complex impedance add additional support for a proposed physical model and its circuit parameters. With this method, the imaginary part of the impedance versus frequency, on the logarithmic scale, are plotted to determine the frequency ranges where the impedance is linear. The linear portions of these plots indicate separate electrochemical processes and distributed time constants, or constant phase element (CPE) behavior, with the slope of the line being the value of the CPE coefficient, α. This is helpful in determining the most accurate fit model with reasonable results. Since the focus is on the placement of the diffusion component in the circuit model, more emphasis will be on the low frequency range, where the diffusion impedance dominates. In this study, the impedance of a lithium-air battery is measured after various cycles, and Nyquist plot and the impedance versus frequency plot are presented. Fitted equivalent circuit models of mixed diffusion and isolated diffusion are fitted to the Nyquist plots and are compared to the results of the graphical analysis technique. Acknowledgment This work was supported by the FREEDM ERC program of the National Science Foundation under award number EEC-08212121. Reference [1] M. E. Orazem, N. Pebere, B. Tribollet, J. Electrochem. Soc., 153, B129 (2006). Figure 1

Effect of Oxidation with Lanthanum Sol–Gel Coating on the 330Cb Carburization at 900 °C

Corrosion and carburization are the main reasons for the degradation of mechanical properties of conveyor units. In order to improve alloy carburization resistance, an adherent oxide scale can act as a carbon diffusion barrier. Oxidation of Alloy 330Cb in air at 900 °C was performed in order to grow a protective chromia scale. Nevertheless, after oxidation some scale spallation occurred during cooling. Then, argon-annealed lanthanum sol–gel coatings were applied in order to improve the oxide scale adherence. This led to a thinner chromia scale. The convoluted scale structure suggested that a mixed diffusion process was induced by lanthanum. After carburization during 22 h at 900 °C under Ar/10 % CH4, the Cr2O3 scale was converted into carbides and it appears that is necessary to have a thick oxide scale to get a good protection.

Diffusion of influenza during the winter of 1889-1890 in Switzerland

The influenza of the winter of 1889-90 was one of the first epidemics to spread all over the world. At the time, several people hypothesized that the railway was one of the main vectors of diffusion of this influenza. This hypothesis was defended in Switzerland especially by Schmid, Chief of the Swiss Office of Health, who collected an impressive body of material about the spread of the epidemic in that country. These data on influenza combined with data about the structure of the railway are used in this paper in order to test the hypothesis of a mixed diffusion process, first between communes interconnected by the railway, and secondly, between those communes and neighbouring communes. An event history analysis model taking into account diffusion effects is proposed and estimated. Results show that the hypothesis is supported if the railway network in Switzerland is not taken as a whole but if a distinction between railway companies is made. Jean-Marie LE GOFF, Lecturer and research scientist, Life Course and Social Inequalities Research Center (LINES), Institute of Social Sciences, University of Lausanne, Vidy Building, 1015 Lausanne, Switzerland. E-mail: jean-marie.legoff@unil.ch .

Diffusion of Oxygen in Thermally Grown Oxide Scales

High temperature reactivity of materials under oxidizing atmospheres is based on the formation of protective oxide scales. The protectiveness is obtained when the thermally grown oxide scales are dense, continuous and adherent to the metallic substrates (even during thermal shocks); as a matter of fact, the growth of the scale has to be governed by diffusion of species across the growing scale. The diffusing species are coming from the substrate (metallic ions) and/or from the oxidizing atmosphere (oxygen ions). The understanding of growth mechanisms can be reached by making two stage oxidation experiments, using oxygen isotopes. The experiment consists in oxidizing first the metallic substrate in 16O2, evacuating the oxygen after the desired time, and replacing it by 18O2. The distribution of oxygen isotopes given by secondary ion mass spectrometry (SIMS) across the oxide scale informs about the growth mechanisms: anionic transport, cations diffusion or mixed diffusion processes. The use of marker experiment is able to determine the oxygen diffusion coefficients within the growing scales. In this case, a thick scale is grown under 16O2, followed by a shorter diffusion experiment in 18O2. The distribution of 18O isotope across the scale by plotting ln (18O intensity) versus x (depth in oxide) allows determining the oxygen effective diffusion coefficient in the scale, according to the classical Fick’s law solution. A more accurate analysis of these profiles can provide information about bulk and grain boundary diffusion of oxygen. The results can be related to kinetics according to Wagner’s theory. The proposed work consists firstly in making a state of the art review about oxygen diffusion in thermally grown oxide scales, and secondly in connecting the so-obtained outcome (effective, bulk and grain boundary diffusion) to kinetics results. The proposed oxides are chromia, alumina and zirconia.

Stochastic targets with mixed diffusion processes and viscosity solutions

Let Z t, z ν be a R d+1 -valued mixed diffusion process controlled by ν with initial condition Z t, z ν ( t)= z. In this paper, we characterize the set of initial conditions such that Z t, z ν can be driven above a given stochastic target at time T by proving that the corresponding value function is a discontinuous viscosity solution of a variational partial differential equation. As applications of our main result, we study two examples: a problem of optimal insurance under self-protection and a problem of option hedging under jumping stochastic volatility where the underlying stock pays a random dividend at a fixed date.

Characterization of a 100 cm2 Class Molten Carbonate Fuel Cell with Current Interruption

A current‐interrupter method has been employed to measure the potential relaxation from output to open‐circuit voltage (OCV). The relaxation process represents the voltage losses of molten‐carbonate fuel cells which are composed of ohmic loss (ηIR), reaction overpotential (ηRe), and Nernst loss (ΔEloss). The responses of current interruption for a 100 cm2 class Li‐Na carbonate fuel cell were measured as functions of temperature, gas utilization, and oxidant gas composition . The single cell showed three different relaxation patterns of time regions during potential decay to OCV; the shortest time region (less than 20 μs) is due to ηIR, an intermittent time region (20 μs to 150 ms) is due to ηRe, and the residual time region is due to ΔEloss. The further analysis of ηRe data is consistent with the argument that the oxygen reduction reaction in the single cell is controlled by a mixed diffusion process of superoxide ion and .

The Philippines insurrection and the 1902–4 cholera epidemic: Part I—Epidemiological diffusion processes in war

This, the first of a two-part paper, presents an historical case study of the impact of war and its aftermath upon the spread of an infectious disease, cholera. The choice of area (Philippine Islands), the time (a year-long period in the immediate aftermath of the Insurrection against US annexation, March 1902–February 1903) and the epidemic (cholera) are conditioned by the detailed reports prepared by the Chief Quarantine Officer for the Philippine Islands, and published weekly in the contemporary USPublic Health Reports. The reports include textual accounts of the progress of the epidemic, and numerical evidence regarding the weekly incidence of cholera in over 440 settlements. This information is used to reconstruct the routes by which cholera diffused through the archipelago and to model the diffusion process on the geographical levels of province, island and nation. It is shown that a spatially contagious component dominated the spread of cholera in many geographical areas and at all spatial levels. In contrast, hierarchical diffusion was generally found to be of only limited significance in the transmission process. These results correlate with the lack of a well-developed urban hierarchy in many parts of the Philippines. Part II of the paper, to be published in Vol. 24, Number 2 (April 1998) of theJournal, goes on to compare the spatial diffusion patterns described here with those found in a subsequent peacetime cholera wave in the Philippines (May 1903–February 1904). Results from both parts of the paper are used to assess the manner in which war can affect diffusion processes and the circumstances under which spread is likely to be dominated by contagious, hierarchical or mixed diffusion processes.

Kinetics of the oxidation of metals ? Influence of self-diffusion of metal

The importance of the self-diffusion of a metal during its oxidation by a gas is treated for the cases of nonstoichiometric oxides having either interstitial cation or cation vacancies. We have established a general relationship for the reaction rate when a mixed diffusion process occurs. From this relationship, we have shown that the pressure dependence can be different, according to whether the rate-determining process is the self-diffusion through the metal or through the product.