xpertsystems/hconc012-sample
收藏资源简介:
HC-ONC-012数据集是一个完全合成的多癌种化疗反应队列样本,包含500名患者的单表数据,涵盖18种癌症类型(如NSCLC、乳腺癌、结直肠癌等)和60多种化疗方案(如FOLFOX、FOLFIRINOX、R-CHOP等)。数据集采用CTCAE v5.0毒性分级标准(15个毒性类别)、RECIST 1.1和iRECIST反应评估,并包括癌症特异性生物标志物面板(如EGFR、HER2、MSI等)、剂量追踪(计划/实际剂量、剂量减少、延迟等)、支持性护理(如G-CSF预防、止吐方案)以及Weibull锚定的生存终点(OS、PFS等)。数据集设计用于化疗结果分析、毒性建模、剂量反应分析和支持性护理研究,无真实患者数据,无PHI或再识别风险。数据格式为CSV,许可证为CC-BY-NC-4.0(样本),商业版本需联系XpertSystems.ai。
HC-ONC-012 dataset is a fully synthetic multi-cancer chemotherapy response cohort sample, comprising a single-table cohort of 500 patients spanning 18 cancer types (e.g., NSCLC, Breast, Colorectal) and over 60 chemotherapy regimens (e.g., FOLFOX, FOLFIRINOX, R-CHOP). It incorporates CTCAE v5.0 toxicity grading across 15 categories, RECIST 1.1 and iRECIST response assessments, cancer-specific biomarker panels (e.g., EGFR, HER2, MSI), dose tracking (planned/actual dose, dose reduction, delays, etc.), supportive care (e.g., G-CSF prophylaxis, antiemetic regimens), and Weibull-anchored survival endpoints (OS, PFS, etc.). The dataset is built for chemotherapy outcomes analytics, toxicity modeling, dose-response analysis, and supportive care research, with no real patient data, PHI, or re-identification risks. Format is CSV, licensed under CC-BY-NC-4.0 for the sample, with commercial licensing available via XpertSystems.ai.




