Shape-informed cardiac mechanics surrogates in data-scarce regimes via geometric encoding and generative augmentation

This record contains the data accompanying the paper Shape-informed cardiac mechanics surrogates in data-scarce regimes via geometric encoding and generative augmentation. The dataset supports the shape model component of the work: a neural implicit shape model of the left ventricle (SDF-SM) combined with a classical PCA shape model (PCA-SM). The archive contains normalised left ventricular meshes for 44 patient geometries together with 1000 synthetic geometries generated by the trained SDF-SM and their latent codes. Related links: Preprint: https://arxiv.org/abs/2602.20306 Code repository: https://gitlab.com/davidecarrara98/shape-informed-cardiac-mechanics-surrogate本数据集配套于论文《基于几何编码与生成式增强的数据稀缺场景下的形状感知心脏力学替代模型》。本数据集支撑该研究的形状模型模块：即结合经典主成分分析形状模型（PCA-SM, Principal Component Analysis Shape Model）的左心室神经隐式形状模型（SDF-SM, Signed Distance Function-based Shape Model）。 该存档包含44例患者解剖结构的标准化左心室网格，以及由训练完成的SDF-SM生成的1000例合成解剖结构及其隐空间编码。 相关链接： 预印本：https://arxiv.org/abs/2602.20306 代码仓库：https://gitlab.com/davidecarrara98/shape-informed-cardiac-mechanics-surrogate