RSNA腹部创伤CT（RATIC）数据集
计算机视觉与模式识别
2024-05-31 v1
摘要
RSNA腹部创伤CT（RATIC）数据集是最大的公开可用、标注了创伤损伤的成人腹部CT研究集合。该数据集包含来自14个国家23个机构的4,274项研究。该数据集通过Kaggle免费提供非商业用途，网址为https://www.kaggle.com/competitions/rsna-2023-abdominal-trauma-detection。该数据集是为RSNA 2023腹部创伤检测竞赛创建的，旨在鼓励开发用于CT扫描中腹部损伤检测的先进机器学习模型。该数据集涵盖多个器官（包括肝脏、脾脏、肾脏、肠道和肠系膜）的创伤损伤检测和分类。标注由美国急诊放射学会（ASER）和腹部放射学会（SAR）的专家放射科医生完成。数据集在多个级别进行标注，包括三个实质器官的损伤存在及损伤分级、活动性外渗和肠道损伤的图像级标注，以及每个潜在受损器官的体素级分割。通过发布该数据集，我们希望促进机器学习和腹部创伤领域的研究与开发，从而改善患者护理和预后。
引用
@article{arxiv.2405.19595,
title = {The RSNA Abdominal Traumatic Injury CT (RATIC) Dataset},
author = {Jeffrey D. Rudie and Hui-Ming Lin and Robyn L. Ball and Sabeena Jalal and Luciano M. Prevedello and Savvas Nicolaou and Brett S. Marinelli and Adam E. Flanders and Kirti Magudia and George Shih and Melissa A. Davis and John Mongan and Peter D. Chang and Ferco H. Berger and Sebastiaan Hermans and Meng Law and Tyler Richards and Jan-Peter Grunz and Andreas Steven Kunz and Shobhit Mathur and Sandro Galea-Soler and Andrew D. Chung and Saif Afat and Chin-Chi Kuo and Layal Aweidah and Ana Villanueva Campos and Arjuna Somasundaram and Felipe Antonio Sanchez Tijmes and Attaporn Jantarangkoon and Leonardo Kayat Bittencourt and Michael Brassil and Ayoub El Hajjami and Hakan Dogan and Muris Becircic and Agrahara G. Bharatkumar and Eduardo Moreno Júdice de Mattos Farina and Dataset Curator Group and Dataset Contributor Group and Dataset Annotator Group and Errol Colak},
journal= {arXiv preprint arXiv:2405.19595},
year = {2024}
}
备注
40 pages, 2 figures, 3 tables