AG2307/pengwin-2026-anatomy-segmentation-pelvic

该数据集包含医学图像数据，每个样本包括图像（image）和对应的标签（label），以及患者ID（patient_id）、原始像素间距（original_spacing）、原始图像形状（original_shape）和原始仿射矩阵（original_affine）等元数据。图像和标签均为多维数组，可能用于分割或分类任务。数据集分为训练集（544个样本）和验证集（34个样本），总大小约6.4GB。This dataset contains medical image data, each sample includes an image and its corresponding label, along with metadata such as patient ID, original spacing, original shape, and original affine matrix. The image and label are multi-dimensional arrays, likely used for segmentation or classification tasks. The dataset is split into training (544 samples) and validation (34 samples) sets, with a total size of approximately 6.4 GB.