胸腔X射线影像数据库--MIMIC-CXR数据探索

Thoracic X-ray Image Database-MIMIC-CXR Data Exploration

临床医生诊断、检查患者病症常需借助超声、X射线、CT、MRI等医学成像技术手段,但查阅医学影像需要专业的知识基础,且阅片任务重,重复性高。随着计算机视觉技术的不断发展和进步,医学图像的自动分析已成为人工智能技术辅助医生诊断、治疗患者的主要方式之一。本文将介绍一个可促进医疗智能化技术发展的大型胸部X射线影像数据库——MIMIC-CXR。该数据库收集了2011~2016年间贝斯以色列女执事医疗中心急诊科的65379例患者的227835项影像学研究,同时包含了针对每项研究的放射学文本报告。本文旨在介绍MIMIC-CXR影像数据库的基本数据内容,以及其转化的MIMIC-CXR-JPG数据库的主要结构,同时探索将深度学习技术应用于MIMIC-CXR影像可能的研究方式。

Clinicians often use ultrasound,X-ray,CT,MRI,and other medical imaging techniques to diagnose and check patient symptoms.However,reading such medical images requires specialized knowledge,and the reading task is heavy and repetitive.With the continuous development and progress of computer vision technology,automatic analysis of medical images has become one of the main methods for artificial intelligence technology to assist doctors in diagnosing and treating patients.This article introduces a large chest X-ray image database MIMIC-CXR,which can promote the development of such technologies.The database collected 227,835 imaging studies of 65,379 patients in the emergency department of Beth Israel Deaconess Medical Center from 2011 to 2016,and included radiology text reports for each study.This article aims to introduce the basic data content of MIMIC-CXR image database and the main structure of the converted MIMIC-CXR-JPG database,and explore possible research methods of applying deep learning technology to MIMIC-CXR images.