Large language models (LLMs) are advancing rapidly in medical NLP, yet Traditional Chinese Medicine (TCM) with its distinctive ontology, terminology, and reasoning patterns requires domain-faithful evaluation. Existing TCM benchmarks are fragmented in coverage and scale and rely on non-unified or generation-heavy scoring that hinders fair comparison. We present the LingLanMiDian (LingLan) benchmark, a large-scale, expert-curated, multi-task suite that unifies evaluation across knowledge recall, multi-hop reasoning, information extraction, and real-world clinical decision-making. LingLan introduces a consistent metric design, a synonym-tolerant protocol for clinical labels, a per-dataset 400-item Hard subset, and a reframing of diagnosis and treatment recommendation into single-choice decision recognition. We conduct comprehensive, zero-shot evaluations on 14 leading open-source and proprietary LLMs, providing a unified perspective on their strengths and limitations in TCM commonsense knowledge understanding, reasoning, and clinical decision support; critically, the evaluation on Hard subset reveals a substantial gap between current models and human experts in TCM-specialized reasoning. By bridging fundamental knowledge and applied reasoning through standardized evaluation, LingLan establishes a unified, quantitative, and extensible foundation for advancing TCM LLMs and domain-specific medical AI research. All evaluation data and code are available at https://github.com/TCMAI-BJTU/LingLan and http://tcmnlp.com.

Objective:With the rapid growth of unstructured clinical narratives in electronic health records (EHRs), clinical named entity recognition (NER) has become a crucial technique for extracting structured medical information. However, traditional supervised models such as CRF and BioClinicalBERT rely on costly manual annotations. Although large language model (LLM)-based zero-shot NER reduces the dependency on labeled data, challenges remain in aligning example selection with task granularity and in effectively integrating prompt design with self-improvement frameworks. Materials and Methods:To address these limitations, we propose OEMA, a novel zero-shot clinical NER framework based on ontology-enhanced multi-agent collaboration. OEMA consists of three core components: (1) a self-annotator that autonomously generates candidate examples; (2) a discriminator that leverages SNOMED CT to filter token-level examples based on clinical relevance; and (3) a predictor that incorporates entity-type descriptions to enhance inference consistency and accuracy. Results:Experimental results on three benchmark datasets, including the real-world I2B2 2010 dataset alongside MTSamples and VAERS, demonstrate that OEMA consistently outperforms existing zero-shot baselines under exact-match evaluation across multiple backbone LLMs (including gpt-3.5, gpt-4.1, and gemini-2.5-flash). Moreover, under relaxed-match criteria, OEMA performs comparably to the supervised BioClinicalBERT model while significantly outperforming the traditional CRF method. Discussion:OEMA integrates ontology-guided reasoning with multi-agent collaboration to address two key challenges in zero-shot clinical NER: granularity mismatch and prompt-self-improvement integration. Ablation studies indicate that ontology-based filtering reduces noise and improves semantic alignment, helping to bridge the style gap between synthetic data and real-world clinical narratives. Conclusion:OEMA advances zero-shot clinical NER and achieves performance approaching supervised models under relaxed-match criteria. Future work will focus on continual learning, open-domain adaptation, negation and assertion status detection, and multilingual generalization to further expand its applicability in clinical NLP.

Background and Objective: Integrating multimodal data, such as pathology images and genomics, is crucial for understanding cancer heterogeneity, personalized treatment complexity, and enhancing survival prediction. However, most current prognostic methods are limited to a single domain of histopathology or genomics, inevitably reducing their potential for accurate patient outcome prediction. Despite advancements in the concurrent analysis of pathology and genomic data, existing approaches inadequately address the intricate intermodal relationships. Methods: This paper introduces the CPathomic method for multimodal data-based survival prediction. By leveraging whole slide pathology images to guide local pathological features, the method effectively mitigates significant intermodal differences through a cross-modal representational contrastive learning module. Furthermore, it facilitates interactive learning between different modalities through cross-modal and gated attention modules. Results: The extensive experiments on five public TCGA datasets demonstrate that CPathomic framework effectively bridges modality gaps, consistently outperforming alternative multimodal survival prediction methods. Conclusion: The model we propose, CPathomic, unveils the potential of contrastive learning and cross-modal attention in the representation and fusion of multimodal data, enhancing the performance of patient survival prediction.

Predicting the interactions between compounds and their potential target proteins is crucial in drug discovery. Existing methods often assume that each compound has an adequate number of target proteins available for model training. However, in practice, the number of target proteins associated with compounds is often limited, making it difficult to gather a sufficient number of training examples. This results in learning bias in the model, leading to poor performance on these compounds. Moreover, this issue is evident in widely used datasets, such as GPCR and kinase, where 77.51% and 12.71% of compounds, respectively, have only one target protein available for model training. However, the issue has not been fully explored, presenting a challenge for model development. In this research, we propose CPOne, a framework designed to address the aforementioned issue. CPOne is a meta-learning based approach for compound-protein interaction prediction in one-shot scenario where each compound has only one target protein available for model training. By utilizing a meta compound learner, CPOne extracts the meta representation of compound from each task. Through fast gradient updates, this representation is quickly adapted to generate a compound-specific representation for the current task, thereby improving performance in one-shot scenario. Through comprehensive experiments, we empirically validate the superiority of CPOne, which demonstrates a promising performance improvement over established methods.

MOTIVATION:Drug repositioning (DR), identifying novel indications for approved drugs, is a cost-effective strategy in drug discovery. Despite numerous proposed DR models, integrating network-based features, differential gene expression, and chemical structures for high-performance DR remains challenging. RESULTS:We propose a comprehensive deep pretraining and fine-tuning framework for DR, termed DrugRepPT. Initially, we design a graph pretraining module employing model-augmented contrastive learning on a vast drug-disease heterogeneous graph to capture nuanced interactions and expression perturbations after intervention. Subsequently, we introduce a fine-tuning module leveraging a graph residual-like convolution network to elucidate intricate interactions between diseases and drugs. Moreover, a Bayesian multiloss approach is introduced to balance the existence and effectiveness of drug treatment effectively. Extensive experiments showcase the efficacy of our framework, with DrugRepPT exhibiting remarkable performance improvements compared to SOTA (state of the arts) baseline methods (improvement 106.13% on Hit@1 and 54.45% on mean reciprocal rank). The reliability of predicted results is further validated through two case studies, i.e. gastritis and fatty liver, via literature validation, network medicine analysis, and docking screening. AVAILABILITY AND IMPLEMENTATION:The code and results are available at https://github.com/2020MEAI/DrugRepPT.

The accurate identification of disease-associated genes is crucial for understanding the molecular mechanisms underlying various diseases. Most current methods focus on constructing biological networks and utilizing machine learning, particularly deep learning, to identify disease genes. However, these methods overlook complex relations among entities in biological knowledge graphs. Such information has been successfully applied in other areas of life science research, demonstrating their effectiveness. Knowledge graph embedding methods can learn the semantic information of different relations within the knowledge graphs. Nonetheless, the performance of existing representation learning techniques, when applied to domain-specific biological data, remains suboptimal. To solve these problems, we construct a biological knowledge graph centered on diseases and genes, and develop an end-to-end knowledge graph completion framework for disease gene prediction using interactional tensor decomposition named KDGene. KDGene incorporates an interaction module that bridges entity and relation embeddings within tensor decomposition, aiming to improve the representation of semantically similar concepts in specific domains and enhance the ability to accurately predict disease genes. Experimental results show that KDGene significantly outperforms state-of-the-art algorithms, whether existing disease gene prediction methods or knowledge graph embedding methods for general domains. Moreover, the comprehensive biological analysis of the predicted results further validates KDGene’s capability to accurately identify new candidate genes. This work proposes a scalable knowledge graph completion framework to identify disease candidate genes, from which the results are promising to provide valuable references for further wet experiments. Data and source codes are available at https://github.com/2020MEAI/KDGene.

Target identification is one of the crucial tasks in drug research and development, as it aids in uncovering the action mechanism of herbs/drugs and discovering new therapeutic targets. Although multiple algorithms of herb target prediction have been proposed, due to the incompleteness of clinical knowledge and the limitation of unsupervised models, accurate identification for herb targets still faces huge challenges of data and models. To address this, we proposed a deep learning-based target prediction framework termed HTINet2, which designed three key modules, namely, traditional Chinese medicine (TCM) and clinical knowledge graph embedding, residual graph representation learning, and supervised target prediction. In the first module, we constructed a large-scale knowledge graph that covers the TCM properties and clinical treatment knowledge of herbs, and designed a component of deep knowledge embedding to learn the deep knowledge embedding of herbs and targets. In the remaining two modules, we designed a residual-like graph convolution network to capture the deep interactions among herbs and targets, and a Bayesian personalized ranking loss to conduct supervised training and target prediction. Finally, we designed comprehensive experiments, of which comparison with baselines indicated the excellent performance of HTINet2 (HR@10 increased by 122.7% and NDCG@10 by 35.7%), ablation experiments illustrated the positive effect of our designed modules of HTINet2, and case study demonstrated the reliability of the predicted targets of Artemisia annua and Coptis chinensis based on the knowledge base, literature, and molecular docking.

Objective To develop and evaluate a fine-tuned large language model (LLM) for traditional Chinese medicine (TCM) prescription recommendation named TCMLLM-PR. Methods First, we constructed an instruction-tuning dataset containing 68654 samples (approximately 10 million tokens) by integrating data from eight sources, including four TCM textbooks, Pharmacopoeia of the People’s Republic of China 2020 (CHP), Chinese Medicine Clinical Cases (CMCC), and hospital clinical records covering lung disease, liver disease, stroke, diabetes, and splenic-stomach disease. Then, we trained TCMLLM-PR using ChatGLM-6B with P-Tuning v2 technology. The evaluation consisted of three aspects: (i) comparison with traditional prescription recommendation models (PTM, TCMPR, and PresRecST); (ii) comparison with TCM-specific LLMs (ShenNong, Huatuo, and HuatuoGPT) and general-domain ChatGPT; (iii) assessment of model migration capability across different disease datasets. We employed precision, recall, and F1 score as evaluation metrics. Results The experiments showed that TCMLLM-PR significantly outperformed baseline models on TCM textbooks and CHP datasets, with F1@10 improvements of 31.80% and 59.48%, respectively. In cross-dataset validation, the model performed best when migrating from TCM textbooks to liver disease dataset, achieving an F1@10 of 0.155 1. Analysis of real-world cases demonstrated that TCMLLM-PR's prescription recommendations most closely matched actual doctors’ prescriptions. Conclusion This study integrated LLMs into TCM prescription recommendations, leveraging a tailored instruction-tuning dataset and developing TCMLLM-PR. This study will publicly release the best model parameters of TCMLLM-PR to promote the development of the decision-making process in TCM practices (https://github.com/2020MEAI/TCMLLM).

Background: The task of relation extraction is a crucial component in the construction of a knowledge graph. However, it often necessitates a significant amount of manual annotation, which can be time-consuming and expensive. Distant supervision, as a technique, seeks to mitigate this challenge by generating a large volume of pseudo-training data at a minimal cost, achieved by mapping triple facts onto the raw text. Objective: The aim of this study is to explore the novelty and potential of the distant supervisionbased relation extraction approach. By leveraging this innovative method, we aim to enhance knowledge reliability and facilitate new knowledge discovery, establishing associations between knowledge from specific biomedical data or existing knowledge graphs and literature. Methods: This study presents a methodology to construct a biomedical knowledge graph employing distant supervision techniques. Through establishing links between knowledge entities and relevant literature sources, we methodically extract and integrate information, thereby expanding and enriching the knowledge graph. This study identified five types of biomedical entities (e.g., diseases, symptoms and genes) and four kinds of relationships. These were linked to PubMed literature and divided into training and testing datasets. To mitigate data noise, the training set underwent preprocessing, while the testing set was manually curated. Results: In our research, we successfully associated 230,698 triples from the existing knowledge graph with relevant literature. Furthermore, we identified additional 205,148 new triples directly sourced from these studies. Conclusion: Our study markedly advances the field of biomedical knowledge graph enrichment, particularly in the context of Traditional Chinese Medicine (TCM). By validating a substantial number of triples through literature associations and uncovering over 200,000 new triples, we have made a significant stride in promoting the development of evidence-based medicine in TCM. The results underscore the potential of using a distant supervision-based relation extraction approach to both validate and expand knowledge bases, contributing to the broader progression of evidence-based practices in the realm of TCM.

Due to the large-scale spread of COVID-19, which has a significant impact on human health and social economy, developing effective antiviral drugs for COVID-19 is vital to saving human lives. Various biomedical associations, e.g., drug-virus and viral protein-host protein interactions, can be used for building biomedical knowledge graphs. Based on these sources, large-scale knowledge reasoning algorithms can be used to predict new links between antiviral drugs and viruses. To utilize the various heterogeneous biomedical associations, we proposed a fusion strategy to integrate the results of two tensor decomposition-based models (i.e., CP-N3 and ComplEx-N3). Sufficient experiments indicated that our method obtained high performance (MRR=0.2328). Compared with CP-N3, the mean reciprocal rank (MRR) is increased by 3.3% and compared with ComplEx-N3, the MRR is increased by 3.5%. Meanwhile, we explored the relationship between the performance and relationship types, which indicated that there is a negative correlation (PCC=0.446, P-value=2.26e-194) between the performance of triples predicted by our method and edge betweenness.

Traditional Chinese medicine (TCM) has played an indispensable role in clinical diagnosis and treatment. Based on a patient's symptom phenotypes, computation-based prescription recommendation methods can recommend personalized TCM prescription using machine learning and artificial intelligence technologies. However, owing to the complexity and individuation of a patient's clinical phenotypes, current prescription recommendation methods cannot obtain good performance. Meanwhile, it is very difficult to conduct effective representation for unrecorded symptom terms in an existing knowledge base. In this study, we proposed a subnetwork-based symptom term mapping method (SSTM) and constructed a SSTM-based TCM prescription recommendation method (termed TCMPR). Our SSTM can extract the subnetwork structure between symptoms from a knowledge network to effectively represent the embedding features of clinical symptom terms (especially the unrecorded terms). The experimental results showed that our method performs better than state-of-the-art methods. In addition, the comprehensive experiments of TCMPR with different hyperparameters (i.e., feature embedding, feature dimension, subnetwork filter threshold, and feature fusion) demonstrate that our method has high performance on TCM prescription recommendation and potentially promote clinical diagnosis and treatment of TCM precision medicine.

Biomedical named entity recognition (BioNER) from clinical texts is a fundamental task for clinical data analysis due to the availability of large volume of electronic medical record data, which are mostly in free text format, in real-world clinical settings. Clinical text data incorporates significant phenotypic medical entities (e.g., symptoms, diseases, and laboratory indexes), which could be used for profiling the clinical characteristics of patients in specific disease conditions (e.g., Coronavirus Disease 2019 (COVID-19)). However, general BioNER approaches mostly rely on coarse-grained annotations of phenotypic entities in benchmark text dataset. Owing to the numerous negation expressions of phenotypic entities (e.g., "no fever," "no cough," and "no hypertension") in clinical texts, this could not feed the subsequent data analysis process with well-prepared structured clinical data. In this paper, we developed Human-machine Cooperative Phenotypic Spectrum Annotation System (http://www.tcmai.org/login, HCPSAS) and constructed a fine-grained Chinese clinical corpus. Thereafter, we proposed a phenotypic named entity recognizer: Phenonizer, which utilized BERT to capture character-level global contextual representation, extracted local contextual features combined with bidirectional long short-term memory, and finally obtained the optimal label sequences through conditional random field. The results on COVID-19 dataset show that Phenonizer outperforms those methods based on Word2Vec with an F1-score of 0.896. By comparing character embeddings from different data, it is found that character embeddings trained by clinical corpora can improve F-score by 0.0103. In addition, we evaluated Phenonizer on two kinds of granular datasets and proved that fine-grained dataset can boost methods' F1-score slightly by about 0.005. Furthermore, the fine-grained dataset enables methods to distinguish between negated symptoms and presented symptoms. Finally, we tested the generalization performance of Phenonizer, achieving a superior F1-score of 0.8389. In summary, together with fine-grained annotated benchmark dataset, Phenonizer proposes a feasible approach to effectively extract symptom information from Chinese clinical texts with acceptable performance.

Molecular property prediction is becoming increasingly important in drug and material discovery, and many research works have demonstrated the great potential of machine learning techniques, especially deep learning. This paper presents our proposed solution for CCKS-2022 task 8, a chemical domain knowledge-aware framework for multi-view molecular property prediction. As a generative self-supervised approach to molecular graph representation learning, the framework is based on Knowledge-guided Pre-training of Graph Transformer (KPGT), which adopts a graph transformer guided by molecular fingerprint and descriptor knowledge. In the fine-tuning stage, combined with practical prediction problems, we fuse functional group information and chemical element knowledge graphs to predict molecular properties. From the perspective of chemical structure, KPGT provides structural information of molecular graphs (especially highlighting chemical bonds), and we further integrate chemical domain knowledge, using functional groups and chemical element knowledge graph, which is the information on physicochemical properties of atoms. From molecular graphs to functional groups, and to atoms, the molecular representation is jointly enhanced by multiple views from coarse to fine. When introducing functional group information and chemical element knowledge graph, we propose a novel BiLSTM-based recurrent module to accumulate domain knowledge. Our framework is able to simultaneously consider molecular graph, functional groups, and atomic physicochemical properties in practical predictions to better predict molecular properties. Finally, without using other external knowledge, the AUC-ROC of the test data reaches 0.88587, ranking second among 140 teams, which validates the performance of our approach.

Chinese medicine (CM) was extensively used to treat COVID-19 in China. We aimed to evaluate the real-world effectiveness of add-on semi-individualized CM during the outbreak. A retrospective cohort of 1788 adult confirmed COVID-19 patients were recruited from 2235 consecutive linked records retrieved from five hospitals in Wuhan during 15 January to 13 March 2020. The mortality of add-on semi-individualized CM users and non-users was compared by inverse probability weighted hazard ratio (HR) and by propensity score matching. Change of biomarkers was compared between groups, and the frequency of CMs used was analyzed. Subgroup analysis was performed to stratify disease severity and dose of CM exposure. The crude mortality was 3.8% in the semi-individualized CM user group and 17.0% among the non-users. Add-on CM was associated with a mortality reduction of 58% (HR = 0.42, 95% CI: 0.23 to 0.77, [Formula: see text] = 0.005) among all COVID-19 cases and 66% (HR = 0.34, 95% CI: 0.15 to 0.76, [Formula: see text] = 0.009) among severe/critical COVID-19 cases demonstrating dose-dependent response, after inversely weighted with propensity score. The result was robust in various stratified, weighted, matched, adjusted and sensitivity analyses. Severe/critical patients that received add-on CM had a trend of stabilized D-dimer level after 3-7 days of admission when compared to baseline. Immunomodulating and anti-asthmatic CMs were most used. Add-on semi-individualized CM was associated with significantly reduced mortality, especially among severe/critical cases. Chinese medicine could be considered as an add-on regimen for trial use.

This study aims to explore the topological regularities of the character network of ancient traditional Chinese medicine (TCM) book. We applied the 2-gram model to construct language networks from ancient TCM books. Each text of the book was separated into sentences and a TCM book was generated as a directed network, in which nodes represent Chinese characters and links represent the sequential associations between Chinese characters in the sentences (the occurrence of identical sequential associations is considered as the weight of this link). We first calculated node degrees, average path lengths, and clustering coefficients of the book networks and explored the basic topological correlations between them. Then, we compared the similarity of network nodes to assess the specificity of TCM concepts in the network. In order to explore the relationship between TCM concepts, we screened TCM concepts and clustered them. Finally, we selected the binary groups whose weights are greater than 10 in Inner Canon of Huangdi (ICH, ) and Treatise on Cold Pathogenic Disease (TCPD, ), hoping to find the core differences of these two ancient TCM books through them. We found that the degree distributions of ancient TCM book networks are consistent with power law distribution. Moreover, the average path lengths of book networks are much smaller than random networks of the same scale; clustering coefficients are higher, which means that ancient book networks have small-world patterns. In addition, the similar TCM concepts are displayed and linked closely, according to the results of cosine similarity comparison and clustering. Furthermore, the core words of Inner Canon of Huangdi and Treatise on Cold Pathogenic Diseases have essential differences, which might indicate the significant differences of language and conceptual patterns between theoretical and clinical books. This study adopts language network approach to investigate the basic conceptual characteristics of ancient TCM book networks, which proposes a useful method to identify the underlying conceptual meanings of particular concepts conceived in TCM theories and clinical operations.

Synonym mapping between phenotype concepts from different terminologies is difficult because terminology databases have been developed largely independently. Existing maps of synonymous phenotype concepts from different terminology databases are highly incomplete, and manually mapping is time consuming and laborious. Therefore, building an automatic method for predictive mapping of synonymous phenotypes is of special importance. We propose a classifier-based phenotype mapping prediction model (CPM) to predict synonymous relationships between phenotype concepts from different terminology databases. The model takes network semantic representations of phenotypes as input and predicts synonymous relationships by training binary classifiers with a voting strategy. We compared the performance of the CPM with a similarity-based phenotype mapping prediction model (SPM), which predicts mapping based on the ranked cosine similarity of candidate mapping concepts. Based on a network representation N2V-TFIDF, with a majority voting strategy method MV, the CPM achieved accuracy of 0.943, which was 15.4% higher than that of the SPM using the cosine similarity method (0.789) and 23.8% higher than that of the SSDTM method (0.724) proposed in our previous work.

Background: Previous studies showed that the effect of antivirals for COVID-19 was promising but varied across patient population, and was modest among severe cases. Chinese Medicine (CM) was extensively used and reported effective in China, awaiting further evidence support. We aimed to evaluate the real-world effectiveness of add-on semi-individualized.Methods: A retrospective total sampling cohort of 1788 adult confirmed COVID-19 patients were recruited from all 2235 consecutive records retrieved from 5 hospitals in Wuhan during15 January to 13 March 2020. Consultation notes, laboratory/imaging investigations, pharmacy and prognosis records were linked by an electronic medical record system and verified by at least 2 researchers independently. The mortality of add-on semi-individualized CM users and non-users was compared by weighted hazard ratios of multivariable Cox regression and by propensity score matching. Change of biomarkers was compared between groups and the frequency of CMs used was analysed. Subgroup analysis was performed to stratify disease severity and dose of CM exposure. Sensitivity analyses were conducted to test the robustness.Findings: The crude mortality was 3.8% in the semi-individualized CM user group and 17.0% among the non-users. Add-on CM was associated with a significant mortality reduction of 58% (HR=0.42, 95%CI: 0.23 to 0.77, p=0.005) and 66% (HR=0.34, 95%CI: 0.15 to 0.76, p=0.009) among all and severe/critical COVID-19 cases with dose-dependent response, after inversely weighted with propensity score calculated by age, gender, history of hypertension, diabetes, coronary artery disease and disease severity. The result was robust in various stratified, weighted, matched, adjusted and sensitivity analyses. Severe/critical patients received add-on CM had a trend of stabilized D-dimer level after 3-7 days of admission compared to baseline.Interpretation: Add-on semi-individualized CM was associated with reduced mortality demonstrating dose-dependent response, especially among severe/critical COVID-19 patients. Chinese medicine could be considered as an add-on regimen for trial use.Funding Statement: This work is partially supported by the National Key Research and Development Program (2017YFC1703506 and 2020YFC0841600). Declaration of Interests: No financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.Ethics Approval Statement: This study was approved by the ethics review board of Hubei Provincial Hospital of Traditional Chinese Medicine (HBZY2020-C01-01). Written consent was waived due to the retrospective nature.

Traditional Chinese Medicine (TCM) has received increasing attention as a complementary approach or alternative to modern medicine. However, experimental methods for identifying novel targets of TCM herbs heavily relied on the current available herb-compound-target relationships. In this work, we present an Herb-Target Interaction Network (HTINet) approach, a novel network integration pipeline for herb-target prediction mainly relying on the symptom related associations. HTINet focuses on capturing the low-dimensional feature vectors for both herbs and proteins by network embedding, which incorporate the topological properties of nodes across multi-layered heterogeneous network, and then performs supervised learning based on these low-dimensional feature representations. HTINet obtains performance improvement over a well-established random walk based herb-target prediction method. Furthermore, we have manually validated several predicted herb-target interactions from independent literatures. These results indicate that HTINet can be used to integrate heterogeneous information to predict novel herb-target interactions.

研究性课题的开展是培养创新创业型人才的重要途径,科研平台在研究性课题中发挥着重要的作用.为了缓解科研平台在实践教学过程中资源不足的矛盾,本文设计了一种基于教研一体化的演示平台,可支持学生开展包括综合实验、创新训练等多种模式的实践项目.该平台运行模式灵活、可扩展性好、资源利用率高、教学成效良好.

The prediction of the ICU patients condition plays an important role in helping doctors make treatment plans, distributing medical resources and assessing medical effects. This paper introduces the research and application advances of the methods used to predicting ICU patients' condition at home and abroad from two fields: clinic and machine learning, including acute physiology and chronic health evaluation ( APACHE) , simplified acute physiology score ( SAPS) , logistic regression, Bayes, artificial neural network, support vector machine (SVM), and Adaboost, analyses the predicting models, results and shortcomings of different methods and looks into the future of the prediction methods of ICU patients condition.