SynthRAD2023

该数据集是为了生成放疗中从锥形束计算机断层扫描（CBCT）到合成CT图像的SynthRAD2023大型挑战赛而设计的。它包含了180位患者的图像，每位患者都有一对经过配准的3D头颈CBCT和CT图像。该数据集的分辨率为1毫米立方，中位图像尺寸为256×256×200。数据集被随机分为140、20和20对，分别用于训练、验证和测试。从训练案例中，提取了28,631对2D切片用于训练目的。该数据集涵盖了180位患者的图像，任务是从锥形束计算机断层扫描中生成合成CT。

This dataset was designed for the large-scale SynthRAD2023 challenge focused on generating synthetic computed tomography (CT) images from cone-beam computed tomography (CBCT) in radiotherapy. It contains images from 180 patients, where each patient has a pair of co-registered 3D head-and-neck CBCT and CT images. The dataset has an isotropic spatial resolution of 1 mm, with a median image size of 256×256×200. It is randomly split into 140, 20, and 20 image pairs for training, validation, and testing, respectively. A total of 28,631 pairs of 2D slices were extracted from the training cases for training purposes. This dataset encompasses images from 180 patients, with the core task of generating synthetic CT images from cone-beam computed tomography scans.

搜集汇总

背景与挑战

背景概述

SynthRAD2023是一个用于放射治疗中CT合成的挑战数据集，包含来自三个医学中心的脑部和骨盆区域的MR/CT和CBCT/CT图像对，总计1080对图像。数据集结构清晰，包含训练、验证和测试集，所有图像均经过预处理和匿名化处理。