Suppr超能文献

医学图像的安全存储与多模态搜索

Safe storage and multi-modal search for medical images.

作者信息

Kommeri Jukka, Niinimäki Marko, Müller Henning

机构信息

Helsinki Institute of Physics, CERN, Switzerland.

出版信息

Stud Health Technol Inform. 2011;169:450-4.

Abstract

Modern hospitals produce enormous amounts of data in all departments, from images, to lab results, medication use, and release letters. Since several years these data are most often produced in digital form, making them accessible for researchers to optimize the outcome of care process and analyze all available data across patients. The Geneva University Hospitals (HUG) are no exception with its daily radiology department's output of over 140'000 images in 2010, with a majority of them being tomographic slices. In this paper we introduce tools for uploading and accessing DICOM images and associated metadata in a secure Grid storage. These data are made available for authorized persons using a Grid security framework, as security is a main problem in secondary use of image data, where images are to be stored outside of the clinical image archive. Our tool combines the security and metadata access of a Grid middleware with the visual search that uses GIFT.

摘要

现代医院的各个科室都会产生海量数据,涵盖图像、实验室检查结果、用药情况及出院小结等。近年来,这些数据大多以数字形式生成,便于研究人员获取,以优化医疗过程的结果,并分析所有患者的可用数据。日内瓦大学医院(HUG)也不例外,其放射科在2010年日均产出超过14万张图像,其中大部分是断层扫描切片。在本文中,我们介绍了用于在安全的网格存储中上传和访问DICOM图像及相关元数据的工具。这些数据通过网格安全框架提供给授权人员,因为安全是图像数据二次使用中的一个主要问题,此时图像要存储在临床图像存档之外。我们的工具将网格中间件的安全性和元数据访问与使用GIFT的视觉搜索相结合。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验