Background: Clinical named entity recognition (CNER), whose goal is to automatically identify clinical entities in electronic medical records (EMRs), is an important research direction of clinical text data mining and information extraction. The promotion of CNER can provide support for clinical decision making and medical knowledge base construction, which could then improve overall medical quality. Compared with English CNER, and due to the complexity of Chinese word segmentation and grammar, Chinese CNER was implemented later and is more challenging. Objective: With the development of distributed representation and deep learning, a series of models have been applied in Chinese CNER. Different from the English version, Chinese CNER is mainly divided into character-based and word-based methods that cannot make comprehensive use of EMR information and cannot solve the problem of ambiguity in word representation. Methods: In this paper, we propose a lattice long short-term memory (LSTM) model combined with a variant contextualized character representation and a conditional random field (CRF) layer for Chinese CNER: the Embeddings from Language Models (ELMo)-lattice-LSTM-CRF model. The lattice LSTM model can effectively utilize the information from characters and words in Chinese EMRs; in addition, the variant ELMo model uses Chinese characters as input instead of the character-encoding layer of the ELMo model, so as to learn domain-specific contextualized character embeddings. Results: We evaluated our method using two Chinese CNER datasets from the China Conference on Knowledge Graph and Semantic Computing (CCKS): the CCKS-2017 CNER dataset and the CCKS-2019 CNER dataset. We obtained F1 scores of 90.13% and 85.02% on the test sets of these two datasets, respectively. Conclusions: Our results show that our proposed method is effective in Chinese CNER. In addition, the results of our experiments show that variant contextualized character representations can significantly improve the performance of the model.

Purpose: To identify the independent risk variables that contribute to the emergence of microalbuminuria(MAU) in type 2 diabetes mellitus(T2DM), to develop two different prediction models, and to show the order of importance of the factors in the better prediction model combined with a SHAP(Shapley Additive exPlanations) plot. Methods: Retrospective analysis of data from 981 patients with T2DM from March 2021 to March 2023. This dataset included socio-demographic characteristics, disease attributes, and clinical biochemical indicators. After preprocessing and variable screening, the dataset was randomly divided into training and testing sets at a 7:3 ratio. To address class imbalance, the Synthetic Minority Oversampling Technique (SMOTE) was applied to balance the training set. Subsequently, prediction models for MAU were constructed using two algorithms: Random Forest and BP neural network. The performance of these models was evaluated using k-fold cross-validation (k = 5), and metrics such as the area under the ROC curve (AUC), accuracy, precision, recall, specificity, and F1 score were utilized for assessment. Results: The final variables selected through multifactorial logistic regression analysis were age, BMI, stroke, diabetic retinopathy(DR), diabetic peripheral vascular disease (DPVD), 25 hydroxyvitamin D (25(OH)D), LDL cholesterol, neutrophil-to-lymphocyte ratio (NLR), and glycated haemoglobin (HbA1c) were used to construct the risk prediction models of Random Forest and BP neural network, respectively, and the Random Forest model demonstrated superior overall performance (AUC = 0.87, Accuracy = 0.80, Precision = 0.79, Recall = 0.84, Specificity = 0.76, F1 Score = 0.81). The SHAP feature matrix plot revealed that HbA1c, NLR, and 25(OH)D were the three most significant factors in predicting the development of MAU in T2DM, with 25(OH)D acting as an independent protective factor. Conclusion: Effective identification of MAU in T2DM, therapeutic strategies for controllable high-risk factors, and prevention or delay of diabetic kidney disease(DKD) can all be achieved with the help of the risk prediction model developed in this study.

Background: With the rapid expansion of biomedical literature, biomedical information extraction has attracted increasing attention from researchers. In particular, relation extraction between 2 entities is a long-term research topic. Objective: This study aimed to perform 2 multiclass relation extraction tasks of Biomedical Natural Language Processing Workshop 2019 Open Shared Tasks: relation extraction of Bacteria-Biotope (BB-rel) task and binary relation extraction of plant seed development (SeeDev-binary) task. In essence, these 2 tasks are aimed at extracting the relation between annotated entity pairs from biomedical texts, which is a challenging problem. Methods: Traditional research methods adopted feature- or kernel-based methods and achieved good performance. For these tasks, we propose a deep learning model based on a combination of several distributed features, such as domain-specific word embedding, part-of-speech embedding, entity-type embedding, distance embedding, and position embedding. The multi-head attention mechanism is used to extract the global semantic features of an entire sentence. Meanwhile, we introduced a dependency-type feature and the shortest dependency path connecting 2 candidate entities in the syntactic dependency graph to enrich the feature representation. Results: Experiments show that our proposed model has excellent performance in biomedical relation extraction, achieving F1 scores of 65.56% and 38.04% on the test sets of the BB-rel and SeeDev-binary tasks. Especially in the SeeDev-binary task, the F1 score of our model is superior to that of other existing models and achieves state-of-the-art performance. Conclusions: We demonstrated that the multi-head attention mechanism can learn relevant syntactic and semantic features in different representation subspaces and different positions to extract comprehensive feature representation. Moreover, syntactic dependency features can improve the performance of the model by learning dependency relation between the entities in biomedical texts.

Objective: To design and optimize a health informatics management platform embedded in an official WeChat public account—a widely used social media and messaging application in China—to support chemotherapy-induced myelosuppression (CIM) management and patient self-management among breast cancer patients undergoing chemotherapy. Methods: Guided by a user-centered design approach, the platform was developed and refined in three stages: needs assessment, expert review and iterative refinement, and pilot feasibility testing. Results: Most patients (88.6%) and clinicians (93.9%) recognized the need for such a platform. In the pilot evaluation, the mean overall MAUQ score was 3.63 ± 0.32, suggesting moderate usability and room for further refinement. During pilot testing, the alert system generated 11 CIM-related alerts and delivered corresponding risk-stratified guidance. Conclusions: The platform may support personalized CIM management by enabling remote monitoring, automated alerts, and risk-stratified guidance. Further studies with larger samples and longer follow-up are needed to evaluate its effectiveness in improving patient self-management and clinical follow-up efficiency.