A false-color image of HeLa cells at a resolution of 0.0515 x 0.0515 µm² depicting the nucleus in cyan, mitochondria in magenta, and lysosomes in green: an example of how to correctly represent data in an image. (Helena Jambor and Christopher Schmied)
Images created by a multitude of high-tech instruments are widely used in scientific research, both as illustrations and as sources of data. Recent advances in the field of optical (or optical) microscopy, in particular, have enabled sensitive, rapid, and high-resolution imaging of various samples, making the use of images in scientific articles more popular than ever.
And yet there are no common standards for publishing images. This leads to major problems in an essential element of the scientific process: reproducibility. Any researcher seeking to replicate the results of a study without complete information about how the keyframes of those results were produced faces an impossible task.
“Many imaging scientists around the world are very concerned about the reproducibility crisis,” says North, senior director of the Bioimaging Resource Center (BIRC) at Rockefeller University. “People publish very little information about how they acquire their images and how they analyze them.”
That’s why an international consortium of experts, including Ved Sharma, image analyst at North and BIRC, recently developed easy-to-follow guidelines for publishing images and image analysis, born from their knowledge collective best practices. These guidelines have recently been published in an open access study in Nature Methods.
The guidelines were assembled as part of a two-year project involving dozens of imaging scientists from QUAREP-LiMi (Quality Assessment and Reproducibility of Instruments and Images in Optical Microscopy), a group that includes 554 members from 39 countries.
They include practical checklists that scientists must follow in order to publish fully understandable and interpretable images.
Each checklist is divided into three levels: minimal, which describes the essential requirements for reproducible images; recommended, which enhances the understandability of the image; and ideal, which includes leading best practices.
For example, the checklist includes standards for formatting, color management, and annotation. Indicating the origin of an insert from an image is minimal; it is recommended to provide intensity scales for grayscale, color and pseudo-color; and annotating image details such as pixel size and exposure time is ideal.
North says: “We advise everyone to post their images in black and white rather than color, as your eye is much more sensitive to detail in monochrome. Many investigators like color images because they are pretty and impressive. But they don’t. I don’t realize they are wasting a lot of information. »
Image analysis checklists cover three different types of workflows: established, new, and machine learning. Citing each step is a minimum requirement of an established workflow, for example, while providing a screen recording or tutorial for a new workflow is ideal.
This is particularly relevant because NIH-funded researchers must now include data management protocols to meet the requirements of the new NIH Data Management Policy, said Nord. “People are like, ‘What are we supposed to write in there?’ » This document gives them these guidelines. »
The fact that scientists have a clearly articulated image analysis workflow is also important for scientific journals; As the primary disseminators of scientific knowledge, journals have a vested interest in ensuring that the articles they publish are transparent about how the results were produced. To this end, the journal editors participated in discussions with the study authors.
Guidelines can only increase accessibility to the data contained in an article, Sharma says. “There is so much information that could be included for each image in an article, but most of the time it is not available, or the reader has to dig deep into the article to find out where the information is in order to make sense of the information. image,” says Sharma. “If scientists started adopting even a minimal standard for publishing images, reproducibility would be much easier for everyone.”
/Public broadcast. This material from the original organization/authors may be timely in nature and edited for clarity, style, and length. Mirage.News takes no institutional position or party, and all views, positions, and conclusions expressed herein are solely those of the author(s).View in full here.
