Rhythm Analysis during Cardiopulmonary Resuscitation Using Convolutional Neural Networks

Chest compressions during cardiopulmonary resuscitation (CPR) induce artifacts in the ECG that may provoque inaccurate rhythm classification by the algorithm of the defibrillator. The objective of this study was to design an algorithm to produce reliable shock/no-shock decisions during CPR using convolutional neural networks (CNN). A total of 3319 ECG segments of 9 s extracted during chest compressions were used, whereof 586 were shockable and 2733 nonshockable. Chest compression artifacts were removed using a Recursive Least Squares (RLS) filter, and the filtered ECG was fed to a CNN classifier with three convolutional blocks and two fully connected layers for the shock/no-shock classification. A 5-fold cross validation architecture was adopted to train/test the algorithm, and the proccess was repeated 100 times to statistically characterize the performance. The proposed architecture was compared to the most accurate algorithms that include handcrafted ECG features and a random forest classifier (baseline model). The median (90% confidence interval) sensitivity, specificity, accuracy and balanced accuracy of the method were 95.8% (94.6–96.8), 96.1% (95.8–96.5), 96.1% (95.7–96.4) and 96.0% (95.5–96.5), respectively. The proposed algorithm outperformed the baseline model by 0.6-points in accuracy. This new approach shows the potential of deep learning methods to provide reliable diagnosis of the cardiac rhythm without interrupting chest compression therapy.     

Application of polymeric drilling fluids for reducing wearing of drill bit

The drill bit or cutting tool is a very important part of the rock drilling operation, it is directly associated with the rate of penetration. Bit wearing has significant involvement of time, labor, and wealth. By identifying the effect of various machine and rock parameters on bit wearing, the optimum working condition can be adopted to achieve high drilling efficiency. In this experimental work, the rate of penetration and parameters of drilling fluids (i.e. type, additives, and concentration, etc.) were optimized to identify their respective impact on the bit wearing of impregnated core diamond bit. The study has been conducted on a laboratory rotary drilling setup. The role of aqueous solutions of carboxymethyl cellulose (CMC), guar gum (GG), and polyacrylamide (PAM) has been examined as polymeric drilling fluid additives on sandstone rock samples and compared with the results of tap water alone as drilling fluid. A significant decrement in bit wearing has been noticed with all additives and the maximum reduction of 70.02% was achieved with CMC fluid additive.     

Generative Adversarial Learning of Protein Tertiary Structures

Protein molecules are inherently dynamic and modulate their interactions with different molecular partners by accessing different tertiary structures under physiological conditions. Elucidating such structures remains challenging. Current momentum in deep learning and the powerful performance of generative adversarial networks (GANs) in complex domains, such as computer vision, inspires us to investigate GANs on their ability to generate physically-realistic protein tertiary structures. The analysis presented here shows that several GAN models fail to capture complex, distal structural patterns present in protein tertiary structures. The study additionally reveals that mechanisms touted as effective in stabilizing the training of a GAN model are not all effective, and that performance based on loss alone may be orthogonal to performance based on the quality of generated datasets. A novel contribution in this study is the demonstration that Wasserstein GAN strikes a good balance and manages to capture both local and distal patterns, thus presenting a first step towards more powerful deep generative models for exploring a possibly very diverse set of structures supporting diverse activities of a protein molecule in the cell.     

基于AdaBoost和深度残差网络的时间序列分类

时间序列分类(TSC)是数据挖掘领域中重要且富有挑战性的问题之一.首先将时间序列数据按照Gramian Angular Summation/Difference Field(GASF/GADF)、Markov Transition Field(MTP)、Recurrence Plot(RP)四种方式编码成图像,然后利用深度残差网络(ResNet)对编码的图像进行分类.为了充分利用四种编码图像的的信息以及提高分类的性能,使用AdaBoost对基分类器进行集成.ResNet在反向传播过程需要保存每一层的激活值,为了减少集成过程的内存消耗,利用可逆残差模块对传统残差模块进行替换.在计算分析阶段,从UCR数据中选取部分数据集进行测试,并将测试结果与当前最优的结果进行对比,实验结果表明所提算法的有效性.     

Time-Dependent Hydrogen Bond Network Formation in Glycerol-Based Deep Eutectic Solvents

Over the last few years, Deep Eutectic Solvents have gained popularity as a novel class of green solvents, due to their feasible synthesis and overall low production costs. The properties of glycerol (Gly)-based Deep Eutectic Solvents are frequently associated with the formation of an extended hydrogen bond network. In this study, two-dimensional Nuclear Magnetic Resonance (NMR) spectroscopy is employed to analyse the effect of glycerol oversaturation of the hydrogen bond acceptor, choline chloride (ChCl) on the structural arrangement of glyceline (molar ratio 1 : 2 ChCl:Gly), selected to represent Gly-based Deep Eutectic Solvents. The rearrangement of glycerol molecules, additionally trapping water molecules inside of isolated clusters, is revealed during a time-resolved analysis, performed in the presence of various fractions of water added to solvent. 200 % oversaturated Deep Eutectic Solvent (1 : 4 ChCl:Gly) is found to be a suitable cryoprotectant candidate, based on the revealed glycerol-water interactions.     

Simulation study on ramp inflow for hybrid autonomous driving

In view of the complexity of the ramp traffic system on the expressway and the diversity of hybrid automatic driving, the safety distance of vehicles and the corresponding car following model are analyzed based on the combination of driver’s personality characteristics and the characteristics of different automatic driving levels of the vehicle–road​ coordination system, and the risk model is established to screen the potentially dangerous vehicles in the ramp merging process. Based on game theory, this paper constructs a ramp coordination and merging control model suitable for hybrid automatic driving. The model takes the safety benefit and efficiency benefit as the objective function, predicts the benefit value under different strategies and makes the optimal strategy under the current conditions. The proposed model is verified by the joint simulation of Python and sumo. the simulation results show that the average number of combined vehicles is 1.91 times higher, the combined position is 50.11% earlier and the average driving time is 28.35% less than that without control strategy and there is obvious advantage when the density is high compared with vehicles without control strategy.     

Polyphenolic Compounds Extracted and Purified from Buddleja Globosa Hope (Buddlejaceae) Leaves Using Natural Deep Eutectic Solvents and Centrifugal Partition Chromatography

Chemical profiling of Buddleja globosa was performed by high-performance liquid chromatography coupled to electrospray ionization (HPLC-DAD-ESI-IT/MS) and quadrupole time-of-flight high-resolution mass spectrometry (HPLC-ESI-QTOF/MS). The identification of 17 main phenolic compounds in B. globosa leaf extracts was achieved. Along with caffeoyl glucoside isomers, caffeoylshikimic acid and several verbascoside derivatives (β-hydroxyverbascoside and β-hydroxyisoverbascoside) were identified. Among flavonoid compounds, the presence of 6-hydroxyluteolin-7-O-glucoside, quercetin-3-O-glucoside, luteolin 7-O-glucoside, apigenin 7-O-glucoside was confirmed. Campneoside I, forsythoside B, lipedoside A and forsythoside A were identified along with verbascoside, isoverbascoside, eukovoside and martynoside. The isolation of two bioactive phenolic compounds verbascoside and forsythoside B from Buddleja globosa (Buddlejaceae) was successfully achieved by centrifugal partition chromatography (CPC). Both compounds were obtained in one-step using optimized CPC methodology with the two-phase solvent system comprising ethyl acetate-n-butanol-ethanol-water (0.25:0.75:0.1:1, v/v). Additionally, eight Natural Deep Eutectic Solvents (NADESs) were tested for the extraction of polyphenols and compared with 80% methanol. The contents of verbascoside and luteolin 7-O-glucoside after extraction with 80% methanol were 26.165 and 3.206 mg/g, respectively. Among the NADESs tested in this study, proline- citric acid (1:1) and choline chloride-1, 2- propanediol (1:2) were the most promising solvents. With these NADES, extraction yields for verbascoside and luteolin 7-O-glucoside were 51.045 and 4.387 mg/g, respectively. Taken together, the results of this study confirm that CPC enabled the fast isolation of bioactive polyphenols from B. globosa. NADESs displayed higher extraction efficiency of phenolic and therefore could be used as an ecofriendly alternative to classic organic solvents.     

Visible Light Cross‐Linking of Gelatin Hydrogels Offers an Enhanced Cell Microenvironment with Improved Light Penetration Depth

In this study, the cyto‐compatibility and cellular functionality of cell‐laden gelatin‐methacryloyl (Gel‐MA) hydrogels fabricated using a set of photo‐initiators which absorb in 400–450 nm of the visible light range are investigated. Gel‐MA hydrogels cross‐linked using ruthenium (Ru) and sodium persulfate (SPS), are characterized to have comparable physico‐mechanical properties as Gel‐MA gels photo‐polymerized using more conventionally adopted photo‐initiators, such as 1‐[4‐(2‐hydroxyethoxy)‐phenyl]‐2‐hydroxy‐2‐methyl‐1‐propan‐1‐one (Irgacure 2959) and lithium phenyl(2,4,6‐trimethylbenzoyl) phosphinate (LAP). It is demonstrated that the Ru/SPS system has a less adverse effect on the viability and metabolic activity of human articular chondrocytes encapsulated in Gel‐MA hydrogels for up to 35 days. Furthermore, cell‐laden constructs cross‐linked using the Ru/SPS system have significantly higher glycosaminoglycan content and re‐differentiation capacity as compared to cells encapsulated using I2959 and LAP. Moreover, the Ru/SPS system offers significantly greater light penetration depth as compared to the I2959 system, allowing thick (10 mm) Gel‐MA hydrogels to be fabricated with homogenous cross‐linking density throughout the construct. These results demonstrate the considerable advantages of the Ru/SPS system over traditional UV polymerizing systems in terms of clinical relevance and practicability for applications such as cell encapsulation, biofabrication, and in situ cross‐linking of injectable hydrogels.     

In Search of the Driving Factor for the Microwave Curing of Epoxy Adhesives and for the Protection of the Base Substrate against Thermal Damage

This study used controlled microwaves to elucidate the response of adhesive components to microwaves and examined the advantages of microwave radiation in curing epoxy adhesives. Curing of adhesives with microwaves proceeded very rapidly, even though each component of the adhesive was not efficiently heated by the microwaves. The reason the adhesive cured rapidly is that microwave heating was enhanced by the electrically charged (ionic) intermediates produced by the curing reaction. In contrast, the cured adhesive displayed lower microwave absorption and lower heating efficiency, suggesting that the cured adhesive stopped heating even if it continued to be exposed to microwaves. This is a definite advantage in the curing of adhesives with microwaves, as, for example, adhesives dropped onto polystyrene could be cured using microwave heating without degrading the polystyrene base substrate.