Development and comparison of machine learning-based models for predicting heart failure after acute myocardial infarction.

AIMS: Heart failure (HF) is one of the common adverse cardiovascular events after acute myocardial infarction (AMI), but the predictive efficacy of numerous machine learning (ML) built models is unclear. This study aimed to build an optimal model to predict the occurrence of HF in AMI patients by comparing seven ML algorithms. METHODS: Cohort 1 included AMI patients from 2018 to 2019 divided into HF and control groups. All first routine test data of the study subjects were collected as the features to be selected for the model, and seven ML algorithms with screenable features were evaluated. Cohort 2 contains AMI patients from 2020 to 2021 to establish an early warning model with external validation. ROC curve and DCA curve to analyze the diagnostic efficacy and clinical benefit of the model respectively. RESULTS: The best performer among the seven ML algorithms was XgBoost, and the features of XgBoost algorithm for troponin I, triglycerides, urine red blood cell count, γ-glutamyl transpeptidase, glucose, urine specific gravity, prothrombin time, prealbumin, and urea were ranked high in importance. The AUC of the HF-Lab9 prediction model built by the XgBoost algorithm was 0.966 and had good clinical benefits. CONCLUSIONS: This study screened the optimal ML algorithm as XgBoost and developed the model HF-Lab9 will improve the accuracy of clinicians in assessing the occurrence of HF after AMI and provide a reference for the selection of subsequent model-building algorithms.

Influence of Alkaline Earth Metal Ions on Structures and Luminescent Properties of NamMnUO2(CO3)3(4-m-2n)- (M = Mg, Ca; m, n = 0-2): Time-Resolved Fluorescence Spectroscopy and Ab Initio Studies.

The luminescence spectra of triscarbonatouranyl complexes were determined by experimental and theoretical methods. Time-resolved laser-induced fluorescence spectroscopy was used to monitor spectra of uranyl and bicarbonate solutions at 0.1 mol kgw-1 ionic strength and pH ca. 8. The concentrations of Mg2+ and Ca2+ in the samples were chosen in order to vary the proportions of the alkaline earth ternary uranyl complexes MgUO2(CO3)32-, CaUO2(CO3)32-, and Ca2UO2(CO3)3. The luminescence spectrum of each complex was determined by decomposition in order to compare it with the simulated spectra of model structures NamMnUO2(CO3)3(4-m-2n)- (M = Mg, Ca; m, n = 0-2) obtained by quantum chemical methods. The density functional theory (DFT) and time-dependent (TD)-DFT methods were used with the PBE0 functional to optimize the structures in the ground and excited states, respectively, including relativistic effects at the spin-free level, and water solvent effects using a continuum polarizable conductor model. The changes in the structural parameters were quantified with respect to the nature and the amount of alkaline earth counterions to explain the luminescence spectra behavior. The first low-lying excited state was successfully computed, together with the vibrational harmonic frequencies. The DFT calculations confirmed that uranyl luminescence originates from electronic transitions from one of the four nonbonding 5f orbitals of uranium to an orbital that has a uranyl-σ (5f, 6d) character mixed with the 2p atomic orbitals of the carbonate oxygens. Additional single-point calculations using the more accurate TD-DFT/CAM-B3LYP allow one to determine the position of the luminescence "hot band" for each structure in the range 467-476 nm and compared fairly well with experimental reports at around 465 nm. The complete luminescence spectra were built from theoretical results with the corresponding assignment of the electronic transitions and vibronic modes involved, mainly the U-Oax stretching mode. The resulting calculated spectra showed a very good agreement with experimental band positions and band spacing attributed to MgUO2(CO3)32-, CaUO2(CO3)32-, and Ca2UO2(CO3)3. The evolution of luminescence intensities with the number of alkaline earth metal ions in the structure was also correctly reproduced.

Experimental and computational evidence of U(VI)-OH-Si(OH)4 complexes under alkaline conditions: Implications for cement systems.

The complexation of uranyl hydroxides with orthosilicic acid was investigated by experimental and theoretical methods. Spectroluminescence titration was performed in a glovebox under argon atmosphere at pH 9.2, 10.5 and 11.5, with [U(VI)] = 10-6 and 5 × 10-6 mol kgw-1. The polymerization effects of silicic acid were minimized by ruling out samples with less than 90 % monomeric silicic acid present, identified via UV-Vis spectrometry using the molybdate blue method. Linear regression analysis based on time-resolved laser-induced fluorescence spectroscopy (TRLFS) results yielded the conditional stepwise formation constants of U(VI)-OH-Si(OH)4 complexes at 0.05 mol kgw-1 NaNO3. The main spectroscopic features - characteristic peak positions and decay-time - are reported for the first time for the UO2(OH)2SiO(OH)3- species observed at pH 9.2 and 10.5 and UO2(OH)2SiO2(OH)22- predominant at pH 11.5. Quantum chemical calculations successfully computed the theoretical luminescence spectrum of the complex UO2(OH)2SiO(OH)3- species, thus underpinning the proposed chemical model for weakly alkaline systems. The conditional stability constants were extrapolated to infinite dilution using the Davies equation, resulting in log10ß°(UO2(OH)2SiO(OH)3-) and log10ß°(UO2(OH)2SiO2(OH)22-). Implications for U(VI) speciation in the presence and absence of competing carbonate are discussed for silicate-rich environments expected in certain repository concepts for nuclear waste disposal.

Cm(III) speciation in the presence of citrate from neutral to hyperalkaline conditions and the effect of calcium.

We investigated the binary Cm-citrate system using time-resolved laser fluorescence spectroscopy (TRLFS), parallel factor analysis (PARAFAC), and quantum chemical calculations. Evidence collectively suggests the stepwise coordination and deprotonation of citrate alcohol groups in Cm-cit complexes with two bound citrate moieties upon increasing pH, which is supported by a bathochromic shift in emission spectra, an observed increase in lifetime measurements, and lower energy minima for citrate alcohol involvement versus hydrolysis of the Cm-citrate species. Our PARAFAC results agree with a 3-component model for the Cm-citrate system and offer pure component decompositions, yielding fraction species across the studied pH range that have a correlated slope = 1 as a function of pH. For the first time, evidence of ternary Ca-Cm-citrate complexes was revealed by TRLFS with increasing calcium concentration at fixed pHm. The formation of these ternary complexes was substantiated with density functional theory (DFT) calculations on simple model systems of the complexes. Shared citrate carboxylate groups between calcium and curium were proposed for all three ternary Ca-Cm-cit complexes based on DFT-determined Ca-O and Cm-O distances. Moreover, we found that the ternary complex with both alcohol groups deprotonated is most stable when it shares both two carboxylate and two alcohol groups between Ca and Cm. The presence of shared functional groups highlights the enhanced stability of these ternary complexes. Additional work is warranted to further constrain the stoichiometry, stability constants and dependence on ionic strength of these complexes for purposes of thermodynamic modeling of repository settings.

Implications of recently derived thermodynamic data and specific ionic interaction theory parameters for (Mg/Ca)nUO2(CO3)3(4-2n)- complexes on the predominance of the Mg2+-Ca2+-UO22+-OH--CO32- systems, and application to natural and legacy-mine waters.

The formation of alkaline earth(II)triscarbonatouranyl(VI) (AenUO2(CO3)3(4-2n)-) species that have been evidenced both in laboratory and in-field studies, is important from slightly acidic pH up to near degraded cementitious in carbonated waters. They are also showing distinctive luminescence properties with a hypsochromic shift relative to UO22+. The conditions of pH, activities of alkaline earth(II) free ions (mostly Mg2+ and Ca2+) and carbonate ions (HCO3-) can be predicted from the thermodynamic functions and constants. The predictive validity of the activity of major alkaline ions (mostly Na+) is determined from the models used to describe the ionic strength comportment of these species, particularly using coefficients from the specific ion interaction theory (SIT). The stability domains of these species are better defined as a function of the activity of the constituents, and applied to natural waters. In this work, using recently obtained complete thermodynamic data and SIT coefficients, we will draw the stability domains of the AenUO2(CO3)3(4-2n)- species in combinations of activities of H+, HCO3-, Mg2+, Ca2+, and Na+ for a wide selection of water compositions from the literature. Water samples were collected near a French mining legacy-site (Site du Bosc, Lodève, France). After determining the major ion compositions, we will verify that the luminescence signal of uranium is in agreement with the predicted speciation in the stability domains.

Machine learning predicts the risk of hemorrhagic transformation of acute cerebral infarction and in-hospital death.

BACKGROUND: The incidence of hemorrhagic transformation (HT) during thrombolysis after acute cerebral infarction (ACI) is very high. We aimed to develop a model to predict the occurrence of HT after ACI and the risk of death after HT. METHODS: Cohort 1 is divided into HT and non-HT groups, to train the model and perform internal validation. All first laboratory test results of study subjects were used as features to be selected for machine learning, and the models built by four machine learning algorithms were compared to screen the best algorithm and model. Following that, the HT group was divided into death and non-death for subgroup analysis. Receiver operating characteristic (ROC) curves etc. to evaluate the model. ACI patients in cohort 2 for external validation. RESULTS: In cohort 1, the HT risk prediction model HT-Lab10 built by the XgBoost algorithm performed the best with AUCROC=0.95 (95% CI, 0.93-0.96). Ten features were included in the model, namely B-type natriuretic peptide precursor, ultrasensitive C-reactive protein, glucose, absolute neutrophil value, myoglobin, uric acid, creatinine, Ca2+, Thrombin time, and carbon dioxide combining power. The model also had the ability to predict death after HT with AUCROC=0.85 (95% CI, 0.78-0.91). The ability of HT-Lab10 to predict the occurrence of HT as well as death after HT was validated in cohort 2. CONCLUSIONS: The model HT-Lab10 built using the XgBoost algorithm showed excellent predictive ability in both the occurrence of HT and the risk of HT death, achieving a model with multiple uses.

Effect of temperature on the complexation of triscarbonatouranyl(VI) with calcium and magnesium in NaCl aqueous solution.

The complex formation of triscarbonatouranyl(VI) UO2(CO3)34- with the alkaline earth metal ions Mg2+ and Ca2+ in 0.10 mol kgw-1 NaCl was studied at variable temperatures: 5-30 °C for Mg2+ and 10-50 °C for Ca2+. Under appropriate conditions, the ternary complexes (MnUO2(CO3)3(4-2n)- with n = 1 for Mg, n = {1; 2} for Ca) were identified by time-resolved laser-induced luminescence spectrometry. Their pure spectral components at 50 °C for CanUO2(CO3)3(4-2n)- and 30 °C for MgUO2(CO3)32- were recovered by multivariate curve resolution alternating least-squares analysis. Approximation models were tested to fit the experimental data-the equilibrium constants of complexation measured at different temperatures-and deduce the thermodynamic functions, i.e., enthalpy, entropy, and heat capacity. The weak influence of temperature on complexation constants induces large uncertainties in terms of thermodynamic functions. Assuming the enthalpy is constant with temperature using the Van't Hoff equation, the first stepwise complexation of UO2(CO3)34- by Ca2+ is estimated to be slightly endothermic, with , while the second stepwise complexation of CaUO2(CO3)32- by Ca2+ with is slightly exothermic, . In contrast to Ca2+, the complexation of UO2(CO3)34- by Mg2+ is slightly exothermic, with . These values are not significantly different from zero inasmuch as the uncertainties are important due to a weak dependence of log10 K° values. The entropic character of the complexation is verified as for the first stepwise complexation of UO2(CO3)34- by Ca2+, for the second stepwise complexation of CaUO2(CO3)32- by Ca2+, and for the complexation of UO2(CO3)34- by Mg2+. The energetics of complexation and sensitivity analysis of the model estimates with temperature are discussed. The uranium speciation in the case of the safety of nuclear waste management, using the present thermodynamic functions, provides support to the assessment of underground nuclear waste repositories.

The determination of the thermodynamic constants of MgUO2(CO3)32- complex in NaClO4 and NaCl media by time-resolved luminescence spectroscopy, and applications in different geochemical contexts.

The formation constants and specific ion interaction coefficients of MgUO2(CO3)32- complex were determined in 0.1 to 1.0 mol kgw-1 NaCl and 0.10 to 2.21 mol kgw-1 NaClO4 media in the framework of the specific ion interaction theory (SIT), by time-resolved laser-induced luminescence spectroscopy. The upper limits of ionic strength were chosen in order to limit luminescence quenching effects generated by high concentrations of Cl- and ClO4- already observed during our earlier studies on CanUO2(CO3)3(4-2n)- complexes (Shang & Reiller, Dalton Trans., 49, 466; Shang et al., Dalton Trans., 49, 15443). The cumulative formation constant determined is , and the specific ion interaction coefficients are ε(MgUO2(CO3)32-, Na+) = 0.19 ± 0.11 kgw mol-1 in NaClO4 and ε(MgUO2(CO3)32-, Na+) = 0.09 ± 0.16 kgw mol-1 in NaCl. Two gratings of 300 and 1800 lines per mm were used to acquire MgUO2(CO3)32- luminescence spectra, where the high-resolution 1800 lines per mm grating detected slight spectral shifts for the principal luminescent bands relative to CanUO2(CO3)3(4-2n)-. The applications of the consistent set of thermodynamic constants and ε values for MnUO2(CO3)3(4-2n)- (M = Mg and Ca) were examined in different geochemical contexts, where Mg over Ca concentration ratio varies to help defining the relative importance of these species.

Determination of formation constants and specific ion interaction coefficients for CanUO2(CO3)3(4-2n)- complexes in NaCl solution by time-resolved laser-induced luminescence spectroscopy.

The formation constants of CaUO2(CO3)32- and Ca2UO2(CO3)3(aq) were determined in NaCl medium at ionic strengths between 0.1 and 1 mol kgw-1 using time-resolved laser-induced luminescence spectroscopy (TRLS). Spectroluminescence titration of UO2(CO3)34- complex by Ca2+ were conducted at atmospheric CO2(g) and varying pH values in order to eliminate the eventual precipitation of both schoepite (UO3 : 2H2O) and calcite (CaCO3) in aqueous solutions. To identify the stoichiometry of calcium, the slope analyses corrected by the Ringböm coefficient for UO2(CO3)34- relative to pH and CO2(g)-instead of typical expression relative to UO22+ and CO32--was applied in this work. Satisfactory linear fits assessed the conditional stepwise formation constants in the range of ionic strength employed in this work, the values of which are in good agreement with literature data at comparable ionic strengths. Extrapolations to infinite dilution were realized in the framework of the specific ion interaction theory (SIT), also providing the evaluation of the specific ion interaction coefficients. The cumulative stability constants at infinite dilution was determined to be log10 ß°(CaUO2(CO3)32-) = 27.20 ± 0.04 and log10 ß°(Ca2UO2(CO3)3(aq)) = 30.49 ± 0.05, which are in good agreement with extrapolation proposed elsewhere in literature using a different extrapolation framework. The specific ion interaction coefficients were found to be ε(CaUO2(CO3)32-,Na+) = (0.29 ± 0.11) and ε(Ca2UO2(CO3)3(aq),NaCl) = (0.66 ± 0.12) kgw mol-1. Integration of alkali metals into the ternary species may explain these positive and relatively large interaction coefficients. Implications on the speciation of uranium in clay groundwaters, representative of radioactive waste repositories, and in seawater are discussed.

Spectroluminescence measurements of the stability constants of CanUO2(CO3)3(4-2n)- complexes in NaClO4 medium and the investigation of interaction effects.

The stability constants of ternary calcium uranyl tricarbonate complexes, CaUO2(CO3)32- and Ca2UO2(CO3)3(aq), were determined in NaClO4 medium at various ionic strengths using time-resolved laser-induced luminescence spectroscopy (TRLS) - also known as time-resolved laser-induced fluorescence spectroscopy (TRLFS). As in a previous study, the potential precipitation of schoepite (UO3·2H2O) and calcite (CaCO3) was avoided via titration of the triscarbonatouranyl complex with Ca2+ at varying pH values. The Ringböm coefficients relative to UO2(CO3)34- were individually evaluated under test sample conditions. Steadily enhanced luminescence intensity and increased decay-times were representative of complexation processes. The stoichiometry of calcium was quantified by slope analysis, and its measured intensity was corrected by using the corresponding Ringböm coefficient. The conditional formation constants, i.e. log10 Kn.1.3, were then assessed after rounding off the slope values to their nearest integers. Cumulative formation constants at infinite dilution log10 ß°n.1.3, and specific ion interaction parameters ε were determined based on the experimental origin and slope values, respectively, over the range of 0.1-2.46 mol kgw-1 NaClO4 using the specific ion interaction theory (SIT) approach. The cumulative stability constants are log10 ß°(CaUO2(CO3)32-) = 27.26 ± 0.04 and log10 ß°(Ca2UO2(CO3)3(aq)) = 30.53 ± 0.06. The specific ion interaction coefficients are estimated to be ε(CaUO2(CO3)32-,Na+) = (0.02 ± 0.04) kgw mol-1 and ε(Ca2UO2(CO3)3(aq),NaClO4) = (0.18 ± 0.07) kgw mol-1. These latter values are different from the ones that were previously obtained in NaCl, and underlying causes are discussed from different aspects. This work provides valuable information to address the interaction effects between Ca-UO2-CO3 species and 1 : 1 type electrolytes. It is suggested that the affinity of the cation in a background electrolyte with CanUO2(CO3)3(4-2n)- (n = {1;2}) has to be taken into consideration at high ionic strengths, especially for globally non-charged species.

The redox behavior of uranium on Beishan granite: Effect of Fe2+ and Fe3+ content.

The Beishan granitic area in Gansu Province is a site with the greatest potential for a repository of high-level radioactive waste (HLW) in China. In this study, the redox behavior of uranium on Beishan granite was investigated at pH values from ~4.4 to ~9.2. Due to the presence of Fe2+-containing fluorannite, results showed that U(VI) was partially reduced by the granites from boreholes 2 (486 m) and 28 (670 m) at a relatively low initial pH whether Na2CO3/NaCl or native groundwater was used as a background electrolyte. Partial oxidation of UO2 was observed when UO2 contacted Beishan granite directly. Therefore, this incomplete reduction of U(VI) was mainly attributed to minor Fe3+ that was either originally contained in the granite or generated during U(VI) reduction. Consequently, aliovalent oxides (e.g., U3O8, U3O7, U4O9, etc.) should be the thermodynamically stable phase in Beishan granite. A mechanism involving the dissolution of Fe2+ from the granite structure followed by interfacial adsorption/reaction was proposed for the U(VI) reduction. This study demonstrates that Beishan granite has a good reducing capacity, which is suitable for the immobilization of redox-sensitive radionuclides. However, potential oxidation of spent fuel by Fe3+ in the granite should also been taken into account.