Aneumo
收藏资源简介:
Aneumo数据集是一个由复旦大学和上海人工智能科学研究院构建的大规模合成数据集,专注于颅内动脉瘤的血流动力学研究。该数据集基于466个真实动脉瘤模型,通过切除和变形操作生成了10,000个合成模型,包含466个无动脉瘤模型和9,534个变形动脉瘤模型。数据集提供了类似医学图像的分割掩码文件,并包含在8个稳态流速下测量的血流动力学参数,如流速、压力和壁面剪切应力。数据集的创建过程包括3D模型变形、分割掩码生成、网格生成和血流动力学模拟等步骤。该数据集旨在解决颅内动脉瘤的病理特征和血流动力学机制的研究问题,支持数据驱动的建模和分析,为动脉瘤的精确模拟和预测提供宝贵资源。数据集的应用领域包括血流动力学建模、优化和预测,帮助探索动脉瘤形态、血流特征与破裂风险之间的关系,从而改进临床诊断和治疗决策。
The Aneumo dataset is a large-scale synthetic dataset developed by Fudan University and the Shanghai Research Institute of Artificial Intelligence, focusing on hemodynamic studies of intracranial aneurysms. This dataset is built upon 466 real aneurysm models, and 10,000 synthetic models are generated via resection and deformation operations, including 466 non-aneurysm models and 9,534 deformed aneurysm models. The dataset provides medical image-like segmentation mask files, and includes hemodynamic parameters measured under 8 steady flow velocities, such as flow velocity, pressure, and wall shear stress. The construction process of the dataset includes steps such as 3D model deformation, segmentation mask generation, mesh generation, and hemodynamic simulation. This dataset aims to address research questions regarding the pathological characteristics and hemodynamic mechanisms of intracranial aneurysms, support data-driven modeling and analysis, and provide valuable resources for accurate simulation and prediction of aneurysms. The application scenarios of the dataset include hemodynamic modeling, optimization, and prediction, assisting in exploring the relationship between aneurysm morphology, blood flow characteristics and rupture risk, thereby improving clinical diagnosis and treatment decisions.

- 1Aneumo: A Large-Scale Comprehensive Synthetic Dataset of Aneurysm Hemodynamics复旦大学人工智能创新与孵化研究院, 上海人工智能科学研究院 · 2025年



