Machine learning-based diagnosis and risk factor analysis of cardiocerebrovascular disease based on KNHANES.

The prevalence of cardiocerebrovascular disease (CVD) is continuously increasing, and it is the leading cause of human death. Since it is difficult for physicians to screen thousands of people, high-accuracy and interpretable methods need to be presented. We developed four machine learning-based CVD classifiers (i.e., multi-layer perceptron, support vector machine, random forest, and light gradient boosting) based on the Korea National Health and Nutrition Examination Survey. We resampled and rebalanced KNHANES data using complex sampling weights such that the rebalanced dataset mimics a uniformly sampled dataset from overall population. For clear risk factor analysis, we removed multicollinearity and CVD-irrelevant variables using VIF-based filtering and the Boruta algorithm. We applied synthetic minority oversampling technique and random undersampling before ML training. We demonstrated that the proposed classifiers achieved excellent performance with AUCs over 0.853. Using Shapley value-based risk factor analysis, we identified that the most significant risk factors of CVD were age, sex, and the prevalence of hypertension. Additionally, we identified that age, hypertension, and BMI were positively correlated with CVD prevalence, while sex (female), alcohol consumption and, monthly income were negative. The results showed that the feature selection and the class balancing technique effectively improve the interpretability of models.

Correction: A p-type multi-wall carbon nanotube/Te nanorod composite with enhanced thermoelectric performance.

[This corrects the article DOI: 10.1039/C7RA13572F.].

MnCo2O4 nanoparticles supported on nitrogen and sulfur co-doped mesoporous carbon spheres as efficient electrocatalysts for oxygen catalytic reactions.

The development of efficient bifunctional electrocatalysts for the oxygen reduction and oxygen evolution reactions is essential to address the challenge of sluggish reaction kinetics. MnCo2O4 nanoparticles supported on nitrogen and sulfur co-doped mesoporous carbon spheres are prepared as non-precious metal electrocatalysts by pyrolysis of thiourea and hydrothermal treatment. The co-doping of nitrogen and sulfur from thiourea into the carbon spheres plays an important role in bimetallic covalent coupling with manganese and cobalt oxides. The as-prepared catalysts exhibit promising catalytic performance of the oxygen reduction reaction compared to commercial platinum catalysts due to the existence of highly active sites. Remarkably, the as-prepared catalysts also exhibit promising catalytic activity of the oxygen evolution reaction, comparable to that of commercial ruthenium oxide in terms of the onset potentials and Tafel slope, and show better durability for both the oxygen reduction reaction and oxygen evolution reaction in an alkaline solution.

Facile fabrication of one-dimensional Te/Cu2Te nanorod composites with improved thermoelectric power factor and low thermal conductivity.

In this study, Te/Cu2Te nanorod composites were synthesized using various properties of Cu2Te, and their thermoelectric properties were investigated. The nanorods were synthesized through a solution phase mixing process, using polyvinylpyrrolidone (PVP). With increasing Cu2Te content, the composites exhibited a reduced Seebeck coefficient and enhanced electrical conductivity. These characteristic changes were due to the high electrical conductivity and low Seebeck coefficient of Cu2Te. The composite containing 30 wt.% of Cu2Te nanorods showed the maximum power factor (524.6 µV/K at room temperature). The two types of nanorods were assembled into a 1D nanostructure, and with this structure, thermal conductivity decreased owing to the strong phonon scattering effect. This nanorod composite had a dramatically improved ZT value of 0.3, which was ~545 times larger than that of pristine Te nanorods.

Surface modification of a BN/ETDS composite with aniline trimer for high thermal conductivity and excellent mechanical properties.

Boron nitride (BN) particles surface-treated with different amounts of aniline trimer (AT) were used to prepare thermally conductive polymer composites with epoxy-terminated dimethylsiloxane (ETDS). For the same weight content of BN, the BN composites surface-treated with AT showed better mechanical strength and thermal conductivity than the pure BN composites. This is because of the intercalation of AT between BN and ETDS, which not only increased the wettability but also provided excellent heat transfer pathways. We determined the optimum surface treatment ratio by varying the amount of AT, and the results are discussed regarding the thermal conductivity, storage modulus, and tensile strength. Finally, we established the optimum AT ratio for BN surface treatment.

A p-type multi-wall carbon nanotube/Te nanorod composite with enhanced thermoelectric performance.

In this study, multi-walled carbon nanotube (MWCNT)/tellurium (Te) nanorod composites with various MWCNT contents are prepared and their thermoelectric properties are investigated. The composite samples are prepared by mixing Te nanorods with surface-treated MWCNTs. Te nanorods are synthesized by solution phase mixing using polyvinylpyrrolidone (PVP). The MWCNTs used in this study are surface-treated with a solution consisting of H2SO4 and HNO3. With increasing MWCNT content, the composite samples exhibit a reduction in the Seebeck coefficient and enhanced electrical conductivity. The maximum power factor of 5.53 µW m K-2 is observed at 2% MWCNT at room temperature. The thermal conductivity of the composite reduced after the introduction of MWCNTs into the Te nanorod matrix; this is attributed to the generation of heterostructured interfaces between MWCNTs and the Te nanorods. At room temperature, the composites containing 2% MWCNTs exhibited the maximum thermoelectric figure of merit (ZT), which is ∼3.91 times larger than that of pure Te nanorods.