The word Crop can be defined as “to trim” or “cut back”.
The Crop tool in most image processing programs is used to trim off the outside edges of a digital image. Cropping can be used to make an image smaller (in pixels) and/or to change the aspect ration (length to width) of the image.
Photographers have historically cropped images to direct the viewer’s eye to a particular subset of a larger image. This can be very useful in art photography, but the same technique used in news or scientific photography has the potential to be misleading.
Example: A tightly cropped image of a political protest can give the impression that there was a large crowd at the event. The reality may be that it was a small group that was crowded together to make it appear in the photo as if there were more people present.
Acquiring scientific images requires us to consider what is important in the sample. The act of selecting a specific area on a sample and the magnification at which to capture the image is a form of cropping. Scientists must be careful to avoid bias, such as selecting images that represent what they think the results of the experiment “should” look like.
Acquiring adequate numbers of representative images of all treatment groups and controls allows the user and their colleagues to carefully review the image data away from the microscope and avoid bias.
Cropping an image for a publication figure is usually considered acceptable.
Scientists should consider their motivation for cropping the image. Is the image being cropped to improve its “composition” or to hide something that disagrees with the lab’s favored hypothesis?
Legitimate reasons for cropping include:
Centering an area of interest
Trimming “empty” space around the edges of an image
Removing a piece of debris from the edge of the image
Questionable forms of cropping would include removing information in a way that changes the context of what remains in the image. Examples:
Cropping out dead or dying cells, leaving only a healthy looking cell
Cropping out gel bands that might disagree with the hypothesis being proposed in the paper
Beware of cropping images too much. Most journals require a minimum of 300 dots per inch (DPI) for images. This means that if an image is to be 3.5 inches wide in the journal, it needs to have 1050 pixels.
Enlarging an image with a small number of pixels can lead to artifacts (see guideline 12).
Do not let image manipulation software replace good science.