ARTICLES

Image Manipulation: Process for Review of Digital Images by Publications Staff

Image manipulation (cropping extraneous information, slightly adjusting image contrast, and resizing an image) is a standard and necessary practice for authors who are preparing figures for publication. However, in light of the rapid advances in digital image technology, including image-editing software, authors now must determine how much image “polishing” is acceptable. Thus, like many basic-science publishers, the journals of the American Physiological Society (APS) now provide guidance to authors about appropriate image manipulation (www.the-aps.org/mm/Publications/Preparing-Your-Manuscript/PreparingFigures). The guidelines were adapted from the Journal of Cell Biology (Rossner and Yamada, Journal of Cell Biology, Volume 166, Number 1, 2004) and emphasize that digital images should be presented as they appear in the original “capture”. That is, editing of the image should not change the content of the data unless the changes are justified and fully declared in the figure and caption.

Evaluation of Digital Images

Several years ago, the editorial art staff of APS began to review all digital images in figures for manipulation. The review occurs after an article is accepted but before it is published online in the noncopyedited “early view” or in final print format. Because only accepted articles are reviewed, corrections do not have to be initiated for manuscripts that may not ultimately be accepted for publication. If an accepted manuscript is determined to have serious problems with figures, its acceptance may be rescinded. It is important to consider at what point in the publication process staff should review figures for manipulation.

Digital images are checked for consistency in brightness and contrast adjustment, composition, and editing. The initial review can be done on the source file by looking at the entire figure and zooming in. Unnatural solid lines and discontinuity between features in an image are often readily detectable with magnification. Image composition can also be assessed by using such software program tools as “touch up object” (Adobe Acrobat) and “Group Objects” or “Ungroup Objects” (Microsoft PowerPoint). Similarly, adjusting the contrast and brightness in the image may help to determine whether all portions were uniformly adjusted and free of selective editing.

In addition to those simple steps, the US Office of Research Integrity has developed a set of forensic tools for Adobe Photoshop that can be downloaded (ori.hhs.gov/forensictools) to analyze the “fingerprint” of digital images, details that are not visible by eye. Although these tools may require a little training, they provide a thorough evaluation of an image’s composition, particularly for identifying unnatural repetitive patterns, duplication, selective editing, and compositing.

Inquiry and Resolution

When a concern is identified, the editor-in-chief and other designated members of the publications staff should discuss the concern and determine whether a query should be initiated. It is important to assume that all items of concern identified during review are non-malicious and will be readily corrected. With that in mind, the publications office sends a letter to the corresponding author, via e-mail, with a figure document that visually details the problem identified. The letter asks the author to review the figures in question as outlined in the figure document and make the necessary corrections to the submitted figures. Original captures are requested if it is thought that the images in question may be composed of multiple pieces or selectively edited.

Usually, authors supply original captures and correct the figures in accordance with the image-manipulation guidelines within a week of the request. Once the original captures and corrected figures are reviewed and approved by the editor-in-chief and other designated publications staff, the manuscript is returned to production. In the rare cases in which the original captures and the figures as submitted do not match, further explanation from all authors of the manuscript may be required before corrections can be considered.

Image-manipulation guidelines that include information about the evaluation and inquiry process help authors prepare figures that meet the standards or correct the ones that do not. They also help journal staff facilitate corrections efficiently and fairly.


CHRISTINA N BENNETT is the publications ethics manager, American Physiological Society, Bethesda, Maryland.