BraTS 2023

该数据集名为BraTS 2023，包含了来自多个地点的脑肿瘤患者的多参数磁共振成像扫描，包括T1、T2、FLAIR和T1Gd等模式，这些扫描经过了去颅骨和刚性配准处理。此外，数据集中提取并裁剪了80个横断面切片，每个切片的尺寸固定为240×240像素。训练集、验证集和测试集分别包含了500、40和40个随机选取的受试者。该数据集的规模涉及多地点的脑肿瘤患者，任务是对脑部磁共振成像进行填补。

The dataset named BraTS 2023 contains multi-parametric magnetic resonance imaging (MRI) scans of brain tumor patients from multiple clinical sites, including T1-weighted, T2-weighted, FLAIR, and T1Gd-weighted sequences. These scans have undergone skull stripping and rigid registration. Additionally, 80 axial cross-sectional slices were extracted and cropped from each scan, with each slice fixed at a dimension of 240 × 240 pixels. The training, validation, and test sets respectively include 500, 40, and 40 randomly selected subjects. This multi-site brain tumor MRI dataset is designed for the task of brain MRI inpainting.