xpertsystems/hc-res-004-sample
收藏资源简介:
HC-RES-004是一个结核病合成数据集样本,由XpertSystems.ai合成数据工厂开发,属于呼吸系统垂直领域的第四个产品。该数据集是一个全球校准的合成结核病队列,覆盖全耐药谱系,包括药物敏感结核病(DS-TB)、多药耐药结核病(MDR-TB)、预广泛耐药结核病(Pre-XDR-TB)、广泛耐药结核病(XDR-TB)和潜伏结核感染(LTBI)。内容涵盖诊断与分子检测(如GeneXpert、LPA、WGS、涂片/培养)、11种药物的耐药性谱(含MIC代理)、治疗方案选择(包括BPaL/BPaLM和STREAM短期方案)、依从性与直接观察治疗、痰菌转换细菌学、世界卫生组织定义的治疗结果、HIV共感染、家庭接触者调查、群体药代动力学/药效学(如NAT2、QTc)以及卫生系统背景。该样本包含500行数据和138列,适用于表格分类、表格回归和生存分析等任务。数据集经过严格验证,符合WHO全球结核病报告等标准,可用于结核病治疗结果预测、药物耐药性分类、依从性建模、HIV-TB共感染管理、接触者调查建模、卫生经济学分析等研究用途,但仅限研究开发,不用于临床决策。
HC-RES-004 is a synthetic tuberculosis (TB) dataset sample developed by XpertSystems.ai’s Synthetic Data Factory, and it is the fourth product in the respiratory disease vertical segment. This dataset constitutes a globally calibrated synthetic TB cohort covering the full spectrum of drug resistance, including drug-susceptible tuberculosis (DS-TB), multidrug-resistant tuberculosis (MDR-TB), pre-extensively drug-resistant tuberculosis (Pre-XDR-TB), extensively drug-resistant tuberculosis (XDR-TB), and latent tuberculosis infection (LTBI). Its content covers diagnostic and molecular testing (e.g., GeneXpert, LPA, WGS, smear/culture), drug resistance profiles for 11 medications (including MIC proxies), treatment regimen selection including BPaL/BPaLM and STREAM short-course regimens, treatment adherence and directly observed therapy (DOT), sputum conversion bacteriology, treatment outcomes as defined by the World Health Organization (WHO), HIV coinfection, household contact investigation, population pharmacokinetics/pharmacodynamics (e.g., NAT2, QTc), and health system context. This sample consists of 500 rows and 138 columns, and is suitable for tasks including tabular classification, tabular regression, and survival analysis. The dataset has been rigorously validated and complies with standards such as the WHO Global Tuberculosis Report. It can be used for research applications including TB treatment outcome prediction, drug resistance classification, adherence modeling, HIV-TB coinfection management, contact investigation modeling, and health economic analysis, but it is for research and development purposes only and shall not be used for clinical decision-making.




