遇见数据集

Sub-acute Ischemic Stroke: MRI scans (T1, T2, DWI, ADC), ground truth and segmentation results

收藏
DataCite Commons2025-05-01 更新2024-07-25 收录
官方服务:

资源简介:

A selection of 37 multi-spectral (T1, T2, DWI and ADC) MRI scans of sub-acute (24 hours to 2 weeks) ischemic stroke cases with two associated ground-truth sets (expert segmentations). Fully anonymized. Ideal for the evaluation of stroke segmentation algorithms. Furthermore contains the a-posteriori and crisp segmentation results of a number of standard classification approaches as described in Maier et al. “Classifiers for Ischemic Stroke Lesion Segmentation: A Comparison Study” PLOS One 2015 (uner review). All processing steps and other details are detailed in the contained README-files. <strong>UPDATE</strong>: A larger, better segmented and pre-processed dataset of ischemic stroke scans is now available under <strong><em>www.isles-challenge.org</em></strong> (ISLES 2015 benchmark at MICCAI 2015) for a fair and direct comparison of ischemic stroke segmentation methods, both sub-acute and acute.

本数据集收录了37例经筛选的亚急性(24小时至2周)缺血性脑卒中患者的多光谱磁共振成像(MRI)扫描结果,涵盖T1、T2、弥散加权成像(DWI)及表观扩散系数(ADC)四种序列。数据集配套两组专家手动分割的金标准分割集,且所有数据均已完成完全匿名化处理,是评估脑卒中分割算法的理想数据集。此外,本数据集还包含若干标准分类方法的后验概率分割结果与硬分割结果,相关方法可参见Maier等人2015年发表于《PLOS One》的论文"Classifiers for Ischemic Stroke Lesion Segmentation: A Comparison Study"(当时处于审稿状态)。所有处理流程与其他细节均已在随附的README文件中详细说明。<strong>更新</strong>:目前已有规模更大、分割与预处理质量更优的缺血性脑卒中扫描数据集发布,可通过<strong><em>www.isles-challenge.org</em></strong>获取(即MICCAI 2015会议上的ISLES 2015基准数据集),用于亚急性与急性缺血性脑卒中分割方法的公平直接对比。

提供机构:
figshare
创建时间:
2016-01-20
二维码
社区交流群
二维码
科研交流群
商业服务