• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>International journal of open source software & processes >A Software for Thorax Images Analysis Based on Deep Learning
【24h】

A Software for Thorax Images Analysis Based on Deep Learning

机译：基于深度学习的胸部图像分析软件

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

People suspected of having COVID-19 need to know quickly if they are infected, so that they can isolate themselves, receive treatment, and inform those with whom they have been in close contact. Currently, the formal diagnosis of COVID-19 infection requires laboratory analysis of blood samples or swabs from the throat and nose. The lab test requires specialized equipment and takes at least 24 hours to produce a result. For this reason, in this paper, the authors tackle the problem of the detection of COVID-19 by developing an open source software to analyze chest x-ray thorax images. The method is based on supervised learning applied to 5000 images. However, deep learning techniques such as convolutional neural network (CNN) and mask R-CNN gives good results comparing with classic medical imaging. Using a dynamic learning rate, they obtained 0.96 accuracy for the training phase and 0.82 for the test. The results of our free tool are comparable to the best state of the art open source systems.
机译：涉嫌Covid-19的人需要很快知道他们是否被感染，因此他们可以隔离自己，接受治疗，并告知那些他们在密切联系的人。目前，Covid-19感染的正式诊断需要从喉咙和鼻子和鼻子的血液样本或拭子的实验室分析。实验室测试需要专门的设备，需要至少24小时才能产生结果。因此，在本文中，作者通过开发开源软件来分析胸X射线胸部图像的开源软件来解决Covid-19的检测问题。该方法基于监督学习应用于5000个图像。然而，诸如卷积神经网络（CNN）和掩模R-CNN等深度学习技术给出了与经典医学成像相比的良好结果。使用动态学习率，它们为训练阶段获得0.96精度和0.82的测试。我们的自由工具的结果与最佳艺术开源系统的最佳状态相当。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号