Dataset: "A Method for Sensitivity Analysis of Automatic Contouring Algorithms Across Different MRI Contrast Weightings Using SyntheticMR"
收藏Figshare2025-01-12 更新2026-04-08 收录
下载链接：
https://figshare.com/articles/dataset/Dataset_A_Method_for_Sensitivity_Analysis_of_Automatic_Contouring_Algorithms_Across_Different_MRI_Contrast_Weightings_Using_SyntheticMR_/28139837/1
下载链接
链接失效反馈官方服务：
资源简介：
For the paper "A Method for Sensitivity Analysis of Automatic Contouring Algorithms Across Different MRI Contrast Weightings Using Synthetic MR" found at: https://doi.org/10.1101/2025.01.10.25319895This dataset contains the ground truth segmentations from three expert annotators (v1, v2, v3) for the left parotid, right parotid, left submandibular, and right submandibular gland. In addition, for each contrast weighting (TR and TE combination), the synthetic image generated from SyMRI as well as the model's predicted automatic contours are also included. This dataset is repeated for a healthy volunteer and a patient. This data follows the following format:<br>Subject-1___healthy-volunteer.zipground_truthParotid_L_v1.niiParotid_L_v2.niiParotid_L_v3.niiParotid_R_v1.nii...modelTR100_TE5structuresParotid_L.nrrdParotid_R.nrrd...img.nrrdTR100_TE10......Subject-2___patient.zipground_truth...model...The contrast weighting combinations (i.e., TR100_TE5) implies a TR = 100 ms and TE = 5 ms.The code to reproduce the analysis from the paper is at: https://github.com/Lucas-Mc/SyntheticMR_DL-sensitivity.
针对收录于https://doi.org/10.1101/2025.01.10.25319895的论文《基于合成磁共振（synthetic MR）的不同磁共振成像（Magnetic Resonance Imaging, MRI）对比度加权下自动勾画算法（automatic contouring algorithms）灵敏度分析方法》，本数据集包含三位专业标注者（v1、v2、v3）针对左侧腮腺、右侧腮腺、左侧下颌下腺及右侧下颌下腺的金标准（ground truth）分割结果。此外，针对每一种对比度加权组合（即重复时间（Repetition Time, TR）与回波时间（Echo Time, TE）的组合），本数据集同时包含由SyMRI生成的合成图像，以及模型预测得到的自动勾画轮廓。本数据集分别针对健康志愿者与患者各构建一套。
本数据集遵循如下格式：
Subject-1___healthy-volunteer.zip
├─ ground_truth
│ ├─ Parotid_L_v1.nii
│ ├─ Parotid_L_v2.nii
│ ├─ Parotid_L_v3.nii
│ ├─ Parotid_R_v1.nii
│ └─ ……
└─ model
├─ TR100_TE5
│ ├─ structures
│ │ ├─ Parotid_L.nrrd
│ │ ├─ Parotid_R.nrrd
│ │ └─ ……
│ └─ img.nrrd
├─ TR100_TE10
└─ ……
Subject-2___patient.zip
├─ ground_truth
│ └─ ……
└─ model
└─ ……
其中对比度加权组合（如TR100_TE5）代表TR=100ms，TE=5ms。本论文中复现相关分析所用的代码可于https://github.com/Lucas-Mc/SyntheticMR_DL-sensitivity获取。
提供机构：
Fuller, Clifton D.; Hwang, Ken-Pin; Ali, Alaa Mohamed Shawky; Floyd, Warren; O'Connell, Nicolette; Thill, Dan; Stancanello, Joseph; Xu, Jiaofeng; Fuentes, David; West, Natalie; Mulder, Samuel; McCullum, Lucas; Wahid, Kareem; Ding, Yao; Belal, Zayne
创建时间：
2025-01-12
