Search

The extraction of medical entities from drug instructions provides fundamental data for the intelligent retrieval of medication information and the construction of medical knowledge graphs, with remarkable research significance and practical value. However, the heterogeneity of medical entities in drug instructions for treating different diseases poses challenges in model training, which requires a large number of annotated samples. To address this issue, a “large model + small model” design approach is used in this research. Specifically, this research proposes a part-label named entity recognition model based on a pre-trained model, which first employs a pre-trained language model fine-tuned on a small number of samples to extract partial entities from drug instructions, and then utilizes a Transformer- based part-label model to further optimize the entity extraction results. The part-label model encodes the input text, identified partial entities, and entity labels using a planar lattice structure, extracts feature representations using Transformer, and predicts entity labels through a conditional random fields (CRF) layer. To reduce the need for annotated training data, a sample data augmentation method is proposed using entity masking strategy on labeled samples to train the part-label model. Experimental results validate the feasibility of the “large model + small model” approach in medical entity extraction, with precision (P), recall (R), and F1 score of 85.0%, 86.1%, and 85.6%, respectively, demonstrating superior performance compared with other learning methods.

Objective To investigate the application value of hydrodissection to expose the recurrent laryngeal nerve in thyroid surgery. Methods Seventy patients undergoing thyroid surgery were randomly divided into the hydrodissection group （n = 35） and control group （n = 35）. In the hydrodissection group， the recurrent laryngeal nerve was exposed by local injection of normal saline. And in the control group， the recurrent laryngeal nerve was exposed by blunt dissection. The operation time， blood loss， length of hospital stay， exposure time of recurrent laryngeal nerve， incidence rates of recurrent laryngeal nerve injury and hypocalcemia were compared between two groups. Results 46 recurrent laryngeal nerves were exposed in the hydrodissection group and 54 in the control group. Surgeons found that hydrodissection could more explicitly display the course of recurrent laryngeal nerve. One case of temporary recurrent laryngeal nerve injury occurred in the hydrodissection group and five cases in the control group， and the difference was not statistically significant （P > 0.05）. No permanent recurrent laryngeal nerve injury was observed in both groups. The exposure time of recurrent laryngeal nerve， operation time， blood loss， incidence rate of hypocalcemia and length of hospital stay did not significantly differ between two groups， and the differences were not statistically significant （all P > 0.05）. Conclusion It is convenient and simple to expose the recurrent laryngeal nerve through hydrodissection. As a safe and reliable Method to expose the recurrent laryngeal nerve， hydrodissection can accurately identify the course of the recurrent laryngeal nerve and avoid mechanical injury induced by excessive dissection of the nerve.

[Objectives] In order to numerically predict the self-propulsion factors of high speed underwater vehicles equipped with Contra-Rotating Propellers (CRP),[Methods] RANS modeling approaches are developed for vehicle resistance,self-propulsion and CRP open-water simulation. In terms of resistance and open-water performance,the modeling accuracy are validated by comparisons with model experiments. Based on quasi-steady and unsteady simulations of a high speed underwater vehicle at full scale,the self-propulsion factors are analyzed and compared.[Results] The comparison of the numerical and experimental results at the model scale indicates that the simulation error margin of vehicle resistance is less than 3%,while those of CRP thrust and torque are less than 2% and 4% respectively. The numerically simulated full-scale resistance is 3% lower than that predicted by the model test data. The self-propulsion factors yielded from full-scale RANS simulations are all reasonable in magnitude. The self-propulsion factors yielded from the quasi-steady and unsteady models differ by less than 2%, indicating that the quasi-steady model is an economical choice for engineering applications.[Conclusions] The present modeling approaches are capable of supplying self-propulsion factors for CRP design with reasonable accuracy,and are expected to enhance design accuracy and work efficiency.

A test system of variable valve timing (VVT) mechanism is designed for the valve phaser of a VVT engine, the working medium flow characteristics of oil control valve (OCV) is measured by test, and the testing principle of VVT dynamic response characteristics is analyzed. The dynamic response curves of the phaser with different OCVs, engine speeds and oil temperatures and pressures are obtained by using the test system. Finally the leakages of phaser are measured. [ABSTRACT FROM AUTHOR]

Objective: The aim of this retrospective study is to evaluate the safety and efficacy of tislelizumab in patients with recurrent/metastatic head and neck squamous cell carcinoma. Methods: Six patients with recurrent/metastatic head and neck squamous cell carcinoma who received tislelizumab monotherapy in our hospital from 2018 to 2020 were retrospectively analyzed. The information of sex, age, TNM stage, efficacy, and adverse reactions were collected. All patients were recruited from the RATIONALE 102 study. The primary end point was the objective response rate, and other end points included progression-free survival and overall survival. We performed tumor immune-related gene sequencing and transcriptome sequencing analysis on the tumor tissues of the patient, and used bioinformatics methods to enrich immune cells and analyze signaling pathways. All analyses were performed using R 4.1. 0 software, SPSS Statistics 24.0 software and GraphPad Prism 8. Results: As of May 31, 2020, the median follow-up time was 26.35 months. The objective response rate with tislelizumab was 50.0%, the median progression-free survival was 6.44 months, and the estimated median survival was 20.07 months. The incidence of grade 3 or higher adverse reactions was 66.7%, including hyponatremia, hypokalemia, hypercalcemia, etc. The expression of macrophage, Treg and neutrophil-related genes are higher in immune-sensitive patients, and the signaling pathways of the intestinal immune network for IgA production, graft versus host disease and autoimmune thyroid disease are significantly activated. Conclusion: Tislelizumab was found to be controllable and tolerable in patients with recurrent/metastatic head and neck squamous cell carcinoma. The response to tislelizumab is related to immune cell infiltration and activation of immune-related signaling pathways., Competing Interests: The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose., (Copyright© by the Editorial Department of Journal of Clinical Otorhinolaryngology Head and Neck Surgery.)