Yaqin Li

Objectives: To predict the occurrence rates and related factors that can lead to Emergence agitation (EA) among pediatric patients after a cleft lip or palate surgery. Materials and Methods: A prospective study was performed to analyze the influencing factors of EA amongst pediatric patients after cleft lip or palate surgeries. The cohort data were collected in the department of oral and maxillofacial surgery of a tertiary stomatological hospital in Chengdu, China, in which patient's demographic variables, monitorization recording of the patient's anesthesia and post-operative situations, and postoperative restlessness assessment records, etc were included. Results: One hundred ninety patients who aged from 3 months to 6 years old, underwent cleft lip/palate repair surgeries and were under general anesthesia participated in this study voluntarily. And this process was conducted from June 2020 to October 2020. The univariate analysis showed that the occurrence of EA was affected by weight, age, fasting duration, operation duration, anesthesia duration, usage of iodoform, and level of pain (p < 0.01). Logistic regression showed that age, fasting duration, operation duration, and level of pain influenced the occurrence of EA (p < 0.05). Conclusions: Age, operation duration, fasting duration, and level of pain could predict the occurrence of EA among postoperative pediatric patients. Clinical Relevance: EA impacts the safety of post-surgery pediatric patients under general anesthesia significantly. Age, operation duration, fasting duration, and level of pain could predict the occurrence of EA among postoperative pediatric patients.

Molecular latent representations, derived from autoencoders (AEs), have been widely used for drug or material discovery over the past couple of years. In particular, a variety of machine learning methods based on latent representations have shown excellent performance on quantitative structure–activity relationship (QSAR) modeling. However, the sequence feature of them has not been considered in most cases. In addition, data scarcity is still the main obstacle for deep learning strategies, especially for bioactivity datasets. In this study, we propose the convolutional recurrent neural network and transfer learning (CRNNTL) method inspired by the applications of polyphonic sound detection and electrocardiogram classification. Our model takes advantage of both convolutional and recurrent neural networks for feature extraction, as well as the data augmentation method. According to QSAR modeling on 27 datasets, CRNNTL can outperform or compete with state-of-art methods in both drug and material properties. In addition, the performances on one isomers-based dataset indicate that its excellent performance results from the improved ability in global feature extraction when the ability of the local one is maintained. Then, the transfer learning results show that CRNNTL can overcome data scarcity when choosing relative source datasets. Finally, the high versatility of our model is shown by using different latent representations as inputs from other types of AEs.