ClarusC64/clinical-hemorrhage-compensation-collapse-v0.1

该数据集用于评估模型是否能够在明显休克症状出现之前检测出由出血引起的崩溃风险。每条数据代表一个患者在一定时间内的多项临床指标变化情况，包括血压、心率、休克指数、血红蛋白、脉压、皮肤灌注等。任务是根据这些指标的变化趋势，分类患者是否处于出血补偿崩溃风险中。数据集还包含了一些干扰变量，如测量噪声和图表噪声。This dataset evaluates whether models can detect hemorrhage-driven collapse before overt shock is visible. Each row represents a short clinical trajectory where blood pressure may remain temporarily preserved while compensatory strain increases. The task is to classify whether the trajectory remains compensated or is moving toward hemorrhage-related collapse. The dataset includes decoy variables such as measurement noise and chart noise.