HD-SIM-MAP: a synthetic dataset with model-based simulations of mean arterial blood pressure changes during hemodialysis

The HD-SIM-MAP dataset is a synthetic (model-based) dataset generated to enable the study of mean arterial blood pressure (MAP) changes during hemodialysis (HD). The dataset includes the profiles of MAP changes during a standard 4-hour HD session simulated using a lumped-parameter, physiologically-based model of the cardiovascular system and the whole-body water and solute kinetics in 5,000 virtual patients with randomly adjusted values of 90 physiological parameters. For each of the 90 selected parameters, a random value was drawn from a normal distribution with the mean equal to the baseline value used originally in the model (with a few exceptions) and the standard deviation (SD) assumed at the level of 10%, 20%, or 40% of the baseline value, depending on the nature of the given parameter and the likelihood of its variation in the population (for some parameters, SD was set below 10% - see Parameters.xlsx). Only values within ±2SD from the mean were accepted. Ultrafiltration was set randomly within ±1 L from the assigned fluid overload. All other parameters as well as dialysis settings were kept constant for all virtual patients (at the levels used in our previous work - see the references below).HD-SIM-MAP数据集是一款基于模型的合成数据集，旨在支撑血液透析（hemodialysis, HD）过程中平均动脉压（mean arterial blood pressure, MAP）变化的相关研究。 该数据集包含5000名虚拟患者在标准4小时血液透析疗程中的平均动脉压变化曲线，这些曲线通过集总参数生理心血管系统模型与全身水-溶质动力学模型模拟得到，其中90项生理参数的取值均经过随机调整。 针对这90项预选参数，均从正态分布中随机抽取取值：分布均值与模型初始基准值保持一致（少数参数除外）；标准差（standard deviation, SD）则根据参数属性及人群中该参数的变异概率，设定为基准值的10%、20%或40%；部分参数的标准差低于10%，详见Parameters.xlsx文件。仅保留均值±2倍标准差范围内的取值。 超滤量设定为在指定液体超负荷量的±1升范围内随机取值。所有其余参数及透析设置在所有虚拟患者中均保持恒定（取值参考本团队此前研究中的设置，详见下文参考文献）。