Bridging Imaging Users to Imaging Analysis - A community survey.
Bridging Imaging Users to Imaging Analysis - A community survey.
复制标题
将影像用户与影像分析联系起来 - 一项社区调查。
DOI：
10.1101/2023.06.05.543701
复制
发表时间：
2023
期刊：
影响因子：
--
通讯作者：
Cimini,BethA
中科院分区：
文献类型：
--
作者：
Sivagurunathan,Suganya;Marcotti,Stefania;Nelson,CarlJ;Jones,MartinL;Barry,DavidJ;Slater,ThomasJA;Eliceiri,KevinW;Cimini,BethA
“桥接成像用户与成像分析”调查由开放生物图像分析中心（COBA）、北美生物成像中心（BINA）和皇家显微学会成像数据分析部（RMS DAIM）于2022年进行，旨在了解成像界的需求。通过多项选择和开放式问题，调查询问了人口统计学，图像分析经验，未来需求以及对工具开发人员和用户角色的建议。调查的参与者来自生命科学和物理科学的不同角色和领域。据我们所知，这是第一次尝试跨社区调查，以弥合物理和生命科学成像之间的知识差距。调查结果表明，受访者的首要需求是文档、图像分析工具使用的详细教程、用户友好的直观软件以及更好的分割解决方案，最好是针对其特定用例定制的格式。工具创建者建议用户熟悉图像分析的基础知识，提供持续的反馈并报告图像分析过程中面临的问题，而用户希望更多的文档和强调工具的友好性。无论计算经验如何，都强烈偏好“书面教程”来获取图像分析知识。我们还观察到，多年来，人们对“办公时间”获得图像分析方法专家意见的兴趣有所增加。结果还显示，在成像社区中，在线讨论论坛在解决图像分析问题方面的使用率低于预期。令人惊讶的是，我们还观察到，尽管人工智能在生物学中的应用越来越多，但受访者对深度/机器学习的兴趣却有所下降。此外，社区建议需要为可用的图像分析工具及其应用程序建立一个公共存储库。社区的意见和建议，在这里全文发布，将有助于图像分析工具创建和教育社区相应地设计和交付资源。
The ‘Bridging Imaging Users to Imaging Analysis’ survey was conducted in 2022 by the Center for Open Bioimage Analysis (COBA), BioImaging North America (BINA) and the Royal Microscopical Society Data Analysis in Imaging Section (RMS DAIM) to understand the needs of the imaging community. Through multichoice and open‐ended questions, the survey inquired about demographics, image analysis experiences, future needs and suggestions on the role of tool developers and users. Participants of the survey were from diverse roles and domains of the life and physical sciences. To our knowledge, this is the first attempt to survey cross‐community to bridge knowledge gaps between physical and life sciences imaging. Survey results indicate that respondents' overarching needs are documentation, detailed tutorials on the usage of image analysis tools, user‐friendly intuitive software, and better solutions for segmentation, ideally in a format tailored to their specific use cases. The tool creators suggested the users familiarise themselves with the fundamentals of image analysis, provide constant feedback and report the issues faced during image analysis while the users would like more documentation and an emphasis on tool friendliness. Regardless of the computational experience, there is a strong preference for ‘written tutorials’ to acquire knowledge on image analysis. We also observed that the interest in having ‘office hours’ to get an expert opinion on their image analysis methods has increased over the years. The results also showed less‐than‐expected usage of online discussion forums in the imaging community for solving image analysis problems. Surprisingly, we also observed a decreased interest among the survey respondents in deep/machine learning despite the increasing adoption of artificial intelligence in biology. In addition, the community suggests the need for a common repository for the available image analysis tools and their applications. The opinions and suggestions of the community, released here in full, will help the image analysis tool creation and education communities to design and deliver the resources accordingly.