摘要
深度学习目前已经广泛应用于医学图像处理领域,利用深度学习对医疗图像进行处理前,需要对来自医院的真实患者数据进行脱敏处理。人工对图像进行脱敏效率低下,为此通过Python设计和实现基于DICOM标准协议的CT图像脱敏批处理系统,并利用PyQt实现脱敏系统的可视化界面效果。分别对500和1000张CT图片进行批量脱敏,实验结果表明,该系统可以在很短时间内有效地去除CT图像中的患者信息,达到脱敏标准。
Deep learning has been widely used in medical image processing. Before using deep learning to process medical images, it is necessary to desensitize the real patient data from hospitals. Given the inefficiency of manual desensitization of images, designs and implements a batch processing system of CT image desensitization in Python, based on DICOM standard. Using PyQt, realizes the interface visualization of this desensitization system. Batch desensitize on 500 and 1000 CT images respectively, experimental results show that the system could effectively remove the patient information in the CT images in a short time and reach the standard of desensitization.
引文
[1]Hinton G E,Salakhutdinov R R.Reducing the Dimensionality of Data with Neural Networks[J].Science,2006,313(5786):504.
[2]中国政府网.建立涉健康医疗数据安全管理制度加强“脱敏”信息和隐私保护的双重研究[EB].http://www.gov.cn/xinwen/2016-06/17/content_5083473.htm.2016,06,17.
[3]陈浩.基于DICOM的远程医疗会诊信息系统的设计与实现[D].西南科技大学,2018,3.18.
[4]陈天莹,陈剑锋.大数据环境下的智能数据脱敏系统[J].通信技术,2016,49(7):915-922.
[5]王鑫,王电钢,母继元,等.基于机器学习的数据脱敏系统研究与设计[J].电力信息与通信技术,2018,16(1):33-38.