Histogram or contrast stretching

Is it valid to perform histogram stretching on an image to be published in a journal as long as I apply the same histogram range to all images I want to compare? If so, is the following statement accurate?
“Contrast enhancement was applied to images using the pixel intensity histogram range of the primary antibody stained image. Equivalent histogram ranges were used for primary antibody images and their matching negative control images.”

I am fairly confident linear adjustments to the image are acceptable in most journals. Gamma correction or other non-linear adjustments are not. Of course, anything taken to extremes that obscures relevant information in the image is always a problem :slight_smile: Live cell imaging would be extremely problematic without linear adjustments!

Many journals have entire sections dedicated to what is acceptable. Nature has an excellent example:
https://www.nature.com/authors/policies/image.html

Microscopy

Authors should be prepared to supply the editors with original data on request, at the resolution collected, from which their images were generated. Cells from multiple fields should not be juxtaposed in a single field; instead multiple supporting fields of cells should be shown as Supplementary Information.

Specific guidelines : Adjustments should be applied to the entire image. Threshold manipulation, expansion or contraction of signal ranges and the altering of high signals should be avoided. If “Pseudo-colouring” and nonlinear adjustment (for example “gamma changes”) are used, this must be disclosed. Adjustments of individual colour channels are sometimes necessary on “merged” images, but this should be noted in the figure legend.

@DianeB,
Linear adjustments are fine, in fact it is not possible to produce image without it, since in many cases the images are collected in the 16-bit color space. The non-linear adjustments are more tricky and whenever used should be reported, I would not use them unless there is a clear rational and understanding of consequences.
Considering your question: “Contrast enhancement was applied to images using the pixel intensity histogram range” - does that mean that you used the same min/max point for contrast stretching for all images? I think it may be not totally clear how particularly the histogram range was used, for example non-linear gamma correction or histogram equalization are also applied over the shown histogram range. Perhaps stating “Linear contrast stretching was applied …” makes it more definite.
Also report (make sure) that the images were collected at the same conditions, i.e. exposure, gain, laser power, etc and thus comparable.

2 Likes