Automated echocardiographic diastolic function grading: A hybrid multi-task deep learning and machine learning approach.

BACKGROUND: Assessing left ventricular diastolic function (LVDF) with echocardiography as per ASE guidelines is tedious and time-consuming. The study aims to develop a fully automatic approach of this procedure by a lightweight hybrid algorithm combining deep learning (DL) and machine learning (ML). METHODS: The model features multi-modality input and multi-task output, measuring LV ejection fraction (LVEF), left atrial end-systolic volume (LAESV), and Doppler parameters: mitral E wave velocity (E), A wave velocity (A), mitral annulus e' velocity (e'), and tricuspid regurgitation velocity (TRmax). The algorithm was trained and tested on two internal datasets (862 and 239 echocardiograms) and validated using three external datasets, including EchoNet-Dynamic and CAMUS. The ASE diastolic function decision tree and total probability theory were used to provide diastolic grading probabilities. RESULTS: The algorithm, named MMnet, demonstrated high accuracy in both test and validation datasets, with Dice coefficients for segmentation between 0.922 and 0.932 and classification accuracies between 0.9977 and 1.0. The mean absolute errors (MAEs) for LVEF and LAESV were 3.7 % and 5.8 ml, respectively, and for LVEF in external validation, MAEs ranged from 4.9 % to 5.6 %. The diastolic function grading accuracy was 0.88 with hard criteria and up to 0.98 with soft criteria which account for the top two probability in total probability theory. CONCLUSIONS: MMnet can automatically grade ASE diastolic function with high accuracy and efficiency by annotating 2D videos and Doppler images.

Removal of Bisphenol A via peroxymonosulfate activation over graphite carbon nitride supported NiCx nanoclusters catalyst: Synergistic oxidation of high-valent nickel-oxo species and singlet oxygen.

In this work, a g-C3N4 supported NiCx nanoclusters catalyst (NiCx-CN) was developed, and its performance in activating peroxymonosulfate (PMS) was evaluated. Mechanism investigation stated that although singlet oxygen (1O2) was formed in the catalytic process, its contribution to BPA elimination was weeny. Interestingly, through the experiment with dimethyl sulfoxide as the probe, it was considered that the high-valent nickel-oxo species (Ni&+=O), generated after the interaction of NiCx-CN and PMS, was the dominating reactive oxygen species (ROS). Theoretical calculations (DFT) implied that NiCx-CN might lose electrons to generate high-valent Ni, which was consistent with the detection of Ni3+ on the surface of the used NiCx-CN. Besides, the prepared NiCx-CN showed advantages in resisting the interference of inorganic anions. Meanwhile, three BPA degradation routes had been proposed based on the transformation intermediates. This study will establish a new protocol for PMS activation using heterogeneous Ni-based catalysts to efficiently degrade organic pollutants via a nonradical mechanism.

CO2-assisted fabrication of silica gel adsorbent in honeycomb rotary wheels for air dehumidification.

A honeycomb rotary wheel fabricated from sheet adsorbent of silica gel is a competitive drying facility for air dehumidification in modern drying and air conditioning industries due to its large contacting area (3,000 m2/m3) and the rapid diffusion of the adsorbate compared to silica gel pellets. The delicate preparation procedure of hygroscopic silica gel is paramount for improved adsorption capacity by optimizing the surface area, pore size, and pore volume of silica gel. In this article, silica gel adsorbent in a honeycomb rotary wheel was fabricated by neutralizing the impregnated water glass solution with a modulus of 3.3 on the glass fiber sheet of the honeycomb matrix using CO2 at different pressure at room temperature instead of corrosive acids. The as-obtained silica gel absorbent was characterized by XRD, scanning electron microscopy (SEM), specific surface area and pore size analysis, and dynamic vapor/gas sorption analysis. The results showed that the as-obtained silica gel adsorbent is uniform in size and tunable in terms of specific surface area, pore size, pore volume, and adsorption capacity by CO2 pressure. The typical silica gel fabricated by CO2 of 0.25 MPa with a specific surface area of 764.86 m2/g, an average micropore size with a diameter of 2.94 nm, and a pore volume of 0.45 ml/g delivers a saturated adsorption capacity of as high as 287.24 mg/g at RH 50%, which is the best in adsorption performance compared to the previously reported results. This provides a new strategy for environment-friendly manufacturing of silica gel adsorbent in honeycomb rotary wheels for air dehumidification.

An Investigation of the Properties of Expanded Polystyrene Concrete with Fibers Based on an Orthogonal Experimental Design.

Expanded polystyrene (EPS) concrete is commonly used as the core material of commercial sandwich panels (CSPs). It is environmentally friendly and lightweight but has poor strength. Adding fibers can improve the microstructure of EPS concrete and reduce the weakening effect of EPS beads on the mechanical properties of concrete. An orthogonal experimental design (OED) was used in this paper to analyze the influence of length and content of polypropylene fiber (PF), glass fiber (GF), and carbon fiber (CF) on the physical and mechanical properties and micromorphology of EPS concrete. Among them, CFs have the most apparent impact on concrete and produce the most significant improvements in all properties. According to the requirements of the flexural performance of CSPs, the splitting tensile strength was taken as the optimization index, and the predicted optimal combination (OC) of EPS concrete with fibers was selected. The variations in the material properties, mechanical properties, and microstructure with age were analyzed. The results show that with increasing age, the dry density, compressive strength, and splitting tensile strength of concrete are markedly improved relative to those of the CSP core material and the control case (CC), and even the degree of hydration is improved.

Extracellular Matrix-receptor Interaction Signaling Genes Associated with Inferior Breast Cancer Survival.

BACKGROUND/AIM: Breast cancer is a common type of cancer in women, and metastasis frequently leads to therapy failure. Using next-generation sequencing (NGS), we aspired to identify the optimal differentially expressed genes (DEGs) for use as prognostic biomarkers for breast cancer. MATERIALS AND METHODS: NGS was used to determine transcriptome profiles in breast cancer tissues and their corresponding adjacent normal tissues from three patients with breast cancer. RESULTS: Herein, 15 DEGs (fold change >4 and <0.25) involved in extracellular matrix (ECM)-receptor interaction signaling were identified through NGS. Among them, our data indicated that high HMMR expression levels were correlated with a poor pathological stage (p<0.001) and large tumor size (p<0.001), whereas high COL6A6 and Reelin (RELN) expression levels were significantly correlated with an early pathological stage (COL6A6: p=0.003 and RELN: p<0.001). Multivariate analysis revealed that high HMMR and SDC1 expression levels were significantly correlated with poor overall survival (OS; HMMR: adjusted hazard ratio [aHR] 1.93, 95% confidence interval [CI]=1.10-3.41, p=0.023; SDC1: [aHR] 2.47, 95%CI=1.28-4.77, p=0.007) for breast cancer. Combined, the effects of HMMR and SDC1 showed a significant correlation with poor OS for patients with breast cancer (high expression for both HMMR and SDC1: [aHR] 3.29, 95%CI=1.52-7.12, p=0.003). CONCLUSION: These findings suggest that HMMR and SDC1 involved in the ECM-receptor interaction signaling pathway could act as effective independent prognostic biomarkers for breast ductal carcinoma.