Clinical Text Analysis
收藏资源简介：
Clinical entity detection, assertion status assignment, and relation extraction are essential in medical text analysis. These techniques enable healthcare professionals, researchers, and medical NLP practitioners to derive valuable insights from clinical literature, electronic health records, and patient notes, enhancing the understanding and management of patient data. This API uses state of the art medical models and is perfect for healthcare data analysts, clinical researchers, and healthcare AI application developers looking to extract detailed and actionable insights from unstructured clinical text.<br/> <br/>Use the provided Streamlit playground application to test this service. <p><br/></p> **Key Features:** - **Entity Recognition** Extracts a wide array of clinical entities using state of the art medical model. - **Assertion Status** Determines the assertion status of each entity and classifies it as hypothetical, past, planned, present, and more. - **Relation Extraction** Links related entities, like drugs with their dosages and frequencies, test results with the tests, and more, which is crucial for building a connected data graph from disjointed text.<br/> **Supported labels:** - **Clinical Entity Labels** Includes categories like Age, Gender, Symptoms, Diseases, Medications, Vital Signs, and many others. - **Assertion Status Labels** Categorizes entities into statuses such as Hypothetical, Past, Present, Planned, and others to provide context. - **Relation Extraction Labels** Identifies relations such as is_finding_of, is_date_of, is_result_of, Drug_BrandName-Dosage, Drug_BrandName-Frequency, Drug_BrandName-Route , Drug_BrandName-Strength, Drug_Ingredient-Dosage, Drug_Ingredient-Frequency, Drug_Ingredient-Route, Drug_Ingredient-Strength.<br/><br/> Benchmarking information :<br/>[Entity Extraction](https://nlp.johnsnowlabs.com/2022/10/19/ner_jsl_en.html#benchmarking) [Assertion Status](https://nlp.johnsnowlabs.com/2021/07/24/assertion_jsl_en.html#benchmarking) [Relation Extraction](https://nlp.johnsnowlabs.com/2021/02/24/re_test_result_date_en.html)
临床实体识别、断言状态分配以及关系抽取是医疗文本分析中的核心任务。此类技术可帮助医疗从业者、科研人员以及医疗自然语言处理（Natural Language Processing, NLP）从业者从临床文献、电子健康档案与患者病历中挖掘有价值的信息，提升对患者数据的理解与管理水平。本API采用前沿医疗模型，专为希望从非结构化临床文本中提取细致且可落地的洞察信息的医疗数据分析人员、临床研究者以及医疗AI应用开发者打造。 可通过提供的Streamlit交互式演示应用测试本服务。 **核心功能：** - **实体识别（Entity Recognition）：** 借助前沿医疗模型提取涵盖多类别的临床实体。 - **断言状态分配（Assertion Status）：** 确定每个实体的断言状态，并将其分类为假设、既往、计划、当前等多种类别。 - **关系抽取（Relation Extraction）：** 关联相关实体，例如药物与其剂量、给药频率，检测结果与其对应检测项目等，这对于从分散文本构建连通的数据图谱至关重要。 **支持的标签类型：** - **临床实体标签（Clinical Entity Labels）：** 涵盖年龄、性别、症状、疾病、药物、生命体征等诸多类别。 - **断言状态标签（Assertion Status Labels）：** 将实体划分为假设、既往、当前、计划等状态，以提供上下文信息。 - **关系抽取标签（Relation Extraction Labels）：** 识别诸如is_finding_of、is_date_of、is_result_of、Drug_BrandName-Dosage、Drug_BrandName-Frequency、Drug_BrandName-Route、Drug_BrandName-Strength、Drug_Ingredient-Dosage、Drug_Ingredient-Frequency、Drug_Ingredient-Route、Drug_Ingredient-Strength等关系。 **基准测试信息：** [Entity Extraction](https://nlp.johnsnowlabs.com/2022/10/19/ner_jsl_en.html#benchmarking) [Assertion Status](https://nlp.johnsnowlabs.com/2021/07/24/assertion_jsl_en.html#benchmarking) [Relation Extraction](https://nlp.johnsnowlabs.com/2021/02/24/re_test_result_date_en.html)
