TFUScapes
收藏资源简介:
TFUScapes数据集是一个大型高分辨率数据集,包含通过解剖学上真实的人类颅骨进行的经颅聚焦超声(tFUS)模拟。数据集由T1加权MRI图像生成,包含了2,500次模拟,覆盖了125个个体,每个个体有20个换能器放置位置。这些模拟使用k-Wave伪谱求解器生成,能够提供毫米级的空间精度,并能针对深部脑结构进行靶向治疗。数据集的创建过程包括将T1加权MRI扫描转换为伪CT体积,并构建模拟配置文件,然后使用GPU加速的k-Wave模拟来生成数据集。该数据集旨在加速计算声学、神经技术和深度学习领域的研究,促进非侵入性脑刺激和治疗的进一步发展。
The TFUScapes dataset is a large-scale high-resolution collection comprising transcranial focused ultrasound (tFUS) simulations performed on anatomically realistic human skulls. Generated from T1-weighted MRI images, the dataset contains 2,500 simulations spanning 125 distinct individuals, with 20 transducer placement positions per individual. These simulations are generated using the k-Wave pseudo-spectral solver, which delivers millimeter-level spatial precision and enables targeted therapy for deep brain structures. The dataset creation process involves converting T1-weighted MRI scans into pseudo-CT volumes, constructing simulation configuration profiles, and generating the dataset via GPU-accelerated k-Wave simulations. This dataset is designed to accelerate research in computational acoustics, neurotechnology, and deep learning, and foster further advancements in non-invasive brain stimulation and therapeutic applications.

- 1A Skull-Adaptive Framework for AI-Based 3D Transcranial Focused Ultrasound Simulation法国斯特拉斯堡大学、法国国家科学研究中心、法国国家卫生与医学研究院、ICube UMR7357、斯特拉斯堡IHU、法国Pessac图像引导治疗公司 · 2025年



