Abstract
新型冠状病毒肺炎(COVID-19)具有高传染性和高致病性,严重威胁人民群众的生命安全和身体健康,快 速准确地检测和诊断 COVID-19 对于疫情控制至关重要。目前 COVID-19 检测诊断方法主要包括核酸检测和基于 医学影像的人工诊断,但是核酸检测耗时较长并且需要专用的测试盒,而基于医学影像的人工诊断过于依赖专业 知识,分析耗时较长且难以发现隐匿病变。随着 X 射线图像和计算机断层扫描图像数据集的相继提出,科研人员 在此基础上构建基于深度学习的 COVID-19 检测诊断模型,有效辅助了医学专家对 COVID-19 的高效诊断治疗。 总结用于 COVID-19 检测诊断的主流影像数据集和相关评价指标,以模型任务和影像数据类型 2 个角度分类介绍 现有基于深度学习的 COVID-19 检测诊断模型,从骨干网络、数据集、影像类型、性能表现、分类种类和开源情况 6 个维度进行比较与分析。此外,介绍用于抗击 COVID-19 的优秀应用系统,并探讨该领域的未来发展趋势。
| Translated title of the contribution | Survey of Studies of COVID-19 Diagnosis Based on Deep Learning |
|---|---|
| Original language | Chinese (Simplified) |
| Article number | 1000-3428(2021)05-0001-15 |
| Number of pages | 15 |
| Journal | Jisuanji Gongcheng/Computer Engineering |
| Volume | 47 |
| Issue number | 5 |
| DOIs | |
| Publication status | Published - May 2021 |
UN SDGs
This output contributes to the following UN Sustainable Development Goals (SDGs)
-
SDG 3 Good Health and Well-being
User-Defined Keywords
- Computer Tomography CT image
- Corona Virus Disease 2019 COVID-19
- Deep learning
- Detection and diagnosis model
- Epidemic control
- X-ray image
Fingerprint
Dive into the research topics of 'Survey of Studies of COVID-19 Diagnosis Based on Deep Learning'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver