Segmentation of bones in medical dual-energy CT volumes using the 3D U-Net convolutional neural network - supplementary data

The data archives contain animated GIFs showing the results of bone segmentation. The algorithm_comparison.rar archive shows the ground truth and results obtained using the JJ2016 and 3D-Unet algorithms. The DECT_comparison.rar archive shows the difference between results obtained from mixed and DECT volumes for the 3D-Unet algorithm. More information is in the report Segmentation of bones in medical dual-energy CT volumes using the 3D U-Net convolutional neural network by José Carlos González Sánchez. The dataset was originally published in DiVA and moved to SND in 2024.本数据档案包含展示骨骼分割结果的GIF动图。algorithm_comparison.rar 压缩包展示了使用JJ2016与3D-Unet算法得到的分割结果与真值（ground truth）。DECT_comparison.rar 压缩包展示了针对3D-Unet算法，混合能量CT体积与双能CT（DECT）体积所得分割结果之间的差异。更多详细信息可参阅José Carlos González Sánchez撰写的《利用3D U-Net卷积神经网络实现医学双能CT体积数据中的骨骼分割》报告。 本数据集最初发布于DiVA平台，并于2024年迁移至SND。

搜集汇总

背景与挑战

背景概述

该数据集是医学双能量CT体积中骨分割研究的补充数据，包含动画GIF文件，用于展示和比较3D U-Net卷积神经网络与其他算法的骨分割结果。数据集发布于2018年，由Linköping University提供，数据可公开访问，主要应用于计算机科学和医学影像领域，支持算法性能评估和影像分析研究。