The Anatomical Tracings of Lesions after Stroke (ATLAS) Dataset - Release 2.0, 2021

To access this data collection, please click on the Restricted Data button above. You will need to download and complete the data use agreement and then email it to [e-mail]. The instructions are in the form. The Anatomical Tracings of Lesions After Stroke (ATLAS) Dataset - Release 2.0 is an open-source data collection consisting a total of 955 T1-weighted MRIs (Magnetic Resonance Imaging) with manually segmented diverse lesions and metadata. ATLAS v2.0 has been split into a public release of 655 T1w MRIs and lesion masks and a hidden test dataset of 300 T1w MRIs. For the hidden dataset, only the T1s MRIs are available. The accompanying manually segmented lesion masks will be made available only for testing algorithm performance in lesion segmentation challenges and competitions. The goal of ATLAS is to provide the research community with a standardized training and testing dataset for lesion segmentation algorithms on T1-weighted MRIs. From 33 cohorts worldwide, 955 MRI images were collected from research groups in the ENIGMA Stroke Recovery Working Group consortium. Images consisted of T1-weighted anatomical MRIs of individuals after stroke. For each MRI, brain lesions were identified and masks were manually drawn on each individual brain in native space using ITK-SNAP (version 3.8.0). After tracing, researchers reviewed and edited lesion masks as necessary using a standardized quality control protocol. In a subset of the data, lesion masks were received from the originating site and edited and checked for quality by the team. All team members received lesion-tracing training and followed a standard operating protocol for tracing lesions to ensure inter-rater reliability on all manually traced masks. All lesion masks were checked twice for quality by trained team members. During the quality control process, researchers ensured that the boundaries of the lesion segmentation were accurate and that all identifiable lesions in the brain were traced. All subject files have undergone a lesion tracing and preprocessing pipeline and are named and stored in accordance with the Brain Imaging Data Structure (BIDS) This dataset is provided in both native subject space and normalized to a standard template (the MNI-152 template).

为访问此数据集，请点击上方的“受限数据”按钮。您需要下载并完成数据使用协议，随后将其电子邮件发送至[e-mail]。具体指导信息已在表格中列出。

《中风后病变解剖追踪》（ATLAS）数据集第二版是一个开源数据集，包含总共955张T1加权磁共振成像（MRI）图像，这些图像经过人工分割，具有多样的病变和元数据。ATLAS v2.0被划分为一个包含655张T1w MRI和病变掩膜的公开版本，以及一个包含300张T1w MRI的隐藏测试数据集。对于隐藏数据集，仅提供T1s MRI。伴随的病变掩膜仅会在病变分割算法性能测试和竞赛中使用。ATLAS的目的是为研究社区提供一个标准化的大语言模型训练和测试数据集，用于T1加权MRI上的病变分割算法。

来自全球33个队列的研究组收集了955张MRI图像，这些图像由ENIGMA中风恢复工作组联盟的研究组提供。图像包括中风后个体的T1加权解剖MRI。对于每张MRI，研究人员确定了大脑病变，并在每个个体的原始空间中手动绘制了大脑的掩膜，使用的是ITK-SNAP（版本3.8.0）。在追踪后，研究人员按照标准化的质量控制协议对病变掩膜进行了审查和编辑。在数据集的一个子集中，病变掩膜由原始站点提供，并由团队进行编辑和质量检查。所有团队成员都接受了病变追踪培训，并遵循了标准操作协议以追踪病变，以确保所有手动追踪的掩膜之间的评分者间可靠性。所有病变掩膜都由经过培训的团队成员进行了两次质量检查。在质量控制过程中，研究人员确保了病变分割边界的准确性，并追踪了大脑中所有可识别的病变。所有主题文件都经过了病变追踪和预处理流程，并按照大脑成像数据结构（BIDS）进行命名和存储。该数据集以原始主题空间和标准化模板（MNI-152模板）进行归一化提供。