TCIA Brain-Tumor-Progression

该集合包括来自20名原发性新诊断为胶质母细胞瘤的受试者的数据集，这些受试者接受了手术和标准的伴随化学放射疗法 (CRT)，然后进行了辅助化疗。每名患者包括两次MRI检查: 在CRT完成后90天内和进展时 (临床确定，并基于临床表现和/或影像学检查结果的组合，并通过治疗或干预的改变来进行)。所有图像集均为DICOM格式，并包含T1w (造影剂前和造影剂后)，FLAIR，T2w，ADC，标准化的脑血流量，标准化的相对脑血容量，标准化的相对脑血容量和二进制肿瘤掩模 (使用T1w图像生成)。灌注图像是在预加载造影剂后由动态敏感性对比 (GRE-EPI DSC) 成像生成的。所有系列都与T1 C图像共同配准。该数据集的目的是评估深度学习算法的性能，以预测肿瘤的进展。

This collection comprises datasets from 20 participants with newly diagnosed primary glioblastoma. All participants underwent surgery followed by standard concomitant chemoradiotherapy (CRT), and subsequently received adjuvant chemotherapy. Each patient had two MRI scans: one performed within 90 days after completion of CRT, and the other at the time of tumor progression, which was clinically confirmed based on a combination of clinical manifestations and/or imaging findings, and validated by changes in treatment or interventions. All image sets are in DICOM format and include T1-weighted (pre- and post-contrast) images, FLAIR, T2-weighted sequences, ADC maps, normalized cerebral blood flow, normalized relative cerebral blood volume, normalized relative cerebral blood volume, and binary tumor masks generated from T1w images. Perfusion images were generated via dynamic susceptibility contrast (GRE-EPI DSC) imaging following pre-administration of contrast agent. All image series were co-registered to the post-contrast T1-weighted (T1 C) images. The purpose of this dataset is to evaluate the performance of deep learning algorithms for predicting tumor progression.

搜集汇总

背景与挑战

背景概述

TCIA Brain-Tumor-Progression数据集包含20名胶质母细胞瘤患者两次MRI检查的DICOM格式数据，含多种成像类型和肿瘤掩模，旨在评估深度学习算法预测肿瘤进展的性能。数据由阿肯色大学医学院于2013年发布。