AG2307/pengwin-2026-fracture-segmentation
收藏资源简介:
该数据集是一个医学图像数据集,主要用于骨骼相关分析,可能涉及骨折检测或分割任务。数据集包含以下特征:image字段(三维浮点数组,表示医学图像数据,如CT扫描)、csm字段(三维无符号整数数组,可能表示分割掩码或分类标签)、case_id(字符串,标识病例)、bone(字符串,表示骨骼类型)、num_fragments(整数,表示碎片数量)、original_spacing(浮点数组,表示原始图像间距)、original_shape(整数数组,表示原始图像形状)和original_affine(浮点数组,表示原始仿射变换矩阵)。数据集分为训练集(940个样本)和验证集(230个样本),总大小约5.26 GB,下载大小约8.43 GB。数据格式支持医学图像处理,适用于计算机视觉和深度学习应用。
This dataset is a medical image dataset primarily designed for bone-related analysis, potentially involving fracture detection or segmentation tasks. It includes the following features: image field (3D float array representing medical image data, such as CT scans), csm field (3D unsigned integer array, likely representing segmentation masks or classification labels), case_id (string identifying the case), bone (string indicating bone type), num_fragments (integer for fragment count), original_spacing (float array for original image spacing), original_shape (integer array for original image shape), and original_affine (float array for original affine transformation matrix). The dataset is split into a training set (940 examples) and a validation set (230 examples), with a total size of approximately 5.26 GB and a download size of about 8.43 GB. The format supports medical image processing and is suitable for computer vision and deep learning applications.




