Many tools are used to predict difficult airway, including bedside screening tests, radiological variables, and ultrasonography. However, the “gold standard” to identify difficult airway before intubation has not been established. The assessment and prediction of difficult airway is receiving increasing attention in clinical practice due to the devastating results of failed oxygenation or intubation. A literature visualization study is necessary to understand the research trend and help tailor future research directions. Science citation index-expanded web of Science database were used to search for literature related to assessment and prediction of difficult airways published before May 9th, 2022. VOS viewer software was used for visual analysis, including literature statistics, and co-occurrence analysis. A total of 2609 articles were included. The amount of relevant research interest and literature is increasing every year. According to co-occurrence network analysis, the research results can be grouped into the following 5 clusters, intubation approaches, intubation in special populations, difficult airway assessment tests, intubation in critical care/emergency settings and education, and laryngoscopes. Co-occurrence overlay analysis showed that video laryngoscopes and index prediction (including computed tomography and ultrasonography), emerged recently and comprised an important percentage of current studies. It can be predicted that future studies should focus on understanding the upper airway anatomy and constructing risk index predictions. Based on current global research trends, risk index predictions are the next hot topics in the evaluation and prediction of difficult airways, and video laryngoscopes will continue to be a hot topic in this field.

No abstract

<p>Machine learning (ML) has brought significant technological innovations in many fields, but it has not been widely embraced by most researchers of natural sciences to date. Traditional understanding and promotion of chemical analysis cannot meet the definition and requirement of big data for running of ML. Over the years, we focused on building a more versatile and low-cost approach to the acquisition of copious amounts of data containing in a chemical reaction. The generated data meet exclusively the thirst of ML when swimming in the vast space of chemical effect. As proof in this study, we carried out a case for acute toxicity test throughout the whole routine, from model building, chip preparation, data collection, and ML training. Such a strategy will probably play an important role in connecting ML with much research in natural science in the future.</p>

Aiming at the mode mixing problem caused by interpolation point selection of conventional EMD (Empirical mode decomposition) method, a secondary iterative sifting EMD method that can avoid mode mixing and achieve high-precision decomposition of HHT (Hilbert–Huang transformation) is proposed based on the theory of EMD. The simulation results show that the proposed method is superior to conventional EMD on the ability to split mixed signal. Finally, the proposed algorithm is applied to the fault diagnosis of rolling bearing and the test results have proved its effectiveness and advantages.

Open-domain dialogue systems have made promising progress in recent years. While the state-of-the-art dialogue agents are built upon large-scale text-based social media data and large pre-trained models, there is no guarantee these agents could also perform well in fast-growing scenarios, such as live streaming, due to the bounded transferability of pretrained models and biased distributions of public datasets from Reddit and Weibo, etc. To improve the essential capability of responding and establish a benchmark in the live opendomain scenario, we introduce the LiveChat dataset, composed of 1.33 million real-life Chinese dialogues with almost 3800 average sessions across 351 personas and fine-grained profiles for each persona. LiveChat is automatically constructed by processing numerous live videos on the Internet and naturally falls within the scope of multi-party conversations, where the issues of Who says What to Whom should be considered. Therefore, we target two critical tasks of response modeling and addressee recognition and propose retrieval-based baselines grounded on advanced techniques. Experimental results have validated the positive effects of leveraging persona profiles and larger average sessions per persona. In addition, we also benchmark the transferability of advanced generation-based models on LiveChat and pose some future directions for current challenges. 1