数据集获取

git clone https://github.com/Fatimaxox/MS-Datasets-Sparks.git

curl -L -o repo.zip https://github.com/Fatimaxox/MS-Datasets-Sparks/archive/refs/heads/main.zip
unzip repo.zip

源站 README 摘录（使用方式）

MS Datasets Repository

Welcome to the MS Datasets Repository! <br>
This repository is dedicated to showcasing datasets used in both segmentation and classification problems related to Multiple Sclerosis (MS). <br>
Here, researchers and individuals interested in MS can easily access various datasets containing MRI images.

MS Segmentation

Dataset 1: Brain MRI Dataset of MS with Consensus Manual Lesion Segmentation and Patient Meta Information

• Source: Baghdad Teaching Hospital
• Number of Subjects: 60 patients
• MRI Sequences Included: T1, T2, and Flair
• Year of Publication: 2022
• Data Type: NIFIT

Dataset 2: ISBI2015 MS Dataset or MS Longitudinal Challenge DataSet

• Source: Johns Hopkins University and Johns Hopkins Hospital
• Number of Subjects: 20 cases
• MRI Sequences Included: T1, T2, MPRAGE, Flair, and PD
• Year of Publication: 2015
• Data Type: NIFIT

Dataset 3: Multiple Sclerosis 2016 Lesion

• Number of Subjects: 53 with 10 cases
• MRI Sequences Included: T1, T2, Flair, and DP
• Year of Publication: 2016
• Data Type: NIFIT

Dataset 4: 2008 MICCAI MS Lesion Segmentation Challenge

• Number of Subjects: 45 with 5 cases
• MRI Sequences Included: T1, T2, Flair
• Last Modified: 2016
• Data Type: NHDR

Dataset 5: MICCAI 2021 - Longitudinal Multiple Sclerosis Lesion Segmentation Challenge

• Source: French Registry on Multiple Sclerosis
• Number of Subjects: 100 with 40 cases
• MRI Sequence Included: Flair
• Year of Publication: 2021
• Data Type: NIFIT

MS Classification

For classification, a local Saudi dataset obtained from King Fahad Specialists Hospital was utilized. A total of 229 patients participated, with MRI orientation covering Brain (Axial, Coronal, Sagittal) and Spinal cord. The MRI sequences included T1, T2, and FLAIR.

Additionally, other resources were utilized:

• Normal cases taken from Brain Tumor MRI Images - 17 Classes with 50 samples taken from T1, T2.
• Not MS cases from Brain Tumor MRI Dataset with 263 samples for the 3 types of tumor.

Contributors

This repository was made by Sparks Team. <br>
MSolutions
🧠🔬📊

数据加载示例（图像类）

from PIL import Image
import glob, os

files = (glob.glob(os.path.join(path, "**", "*.png"), recursive=True)
       + glob.glob(os.path.join(path, "**", "*.jpg"), recursive=True)
       + glob.glob(os.path.join(path, "**", "*.tif"), recursive=True))
print("图像文件数:", len(files))
img = Image.open(files[0]); print("尺寸/模式:", img.size, img.mode)

# torchvision Dataset 方式:
# from torchvision import datasets
# ds = datasets.ImageFolder(path)  # 要求 子目录=类别

目录组织与标注格式以源站说明和下载后实际文件为准。

完整仓库：github.com/Fatimaxox/MS-Datasets-Sparks