Case publishing guidelines
Citation, DOI and article data
Case publishing guidelines are here to help get an idea of the minimum set of expectations that we, as an editorial board, think are acceptable when uploading a case.
A great way to get to grips with the process of uploading cases is to check out our Creating Cases Learning Pathway.
As Radiopaedia.org gets larger and attracts a greater number of people from across the Web, we will need to ensure that cases are as good as possible; these have evolved over the years, and as such, some older cases that you find on the site would no longer fulfill these guidelines.
Please use these guidelines as a checklist prior to upload.
If cases do not meet these criteria, they will not be published. Instead, our editors will push these cases back to draft mode and ask you to either improve the case or keep it unlisted. In some instances (e.g. where there are patient confidentiality concerns) the case may need to be deleted.
It is important that you have permission to upload the case from your institution. This will depend very much on your institution's departmental guidelines, and this varies a great deal from place to place.
Generally, contributing a case on Radiopaedia.org should be considered equivalent to publishing a case in a journal, and as is the case with most journals, written signed consent from the imaged individual is not required, provided the images are appropriately anonymised (see below) and patient confidentiality is in no way compromised.
Many (most?) hospitals have guidelines that allow publication of images provided they are appropriately de-identified, in line with their own privacy policies, without individual consent. Others require individual consent. Still others have a blanket "no" approach. It is, therefore, your responsibility to establish that you are permitted to upload images.
It is also worth considering if any of your colleagues will feel that they have a 'claim' to the images. This again will vary from place to place, as well as among individuals. You are welcome to add a "case courtesy of Dr X" in the case discussion section of the case.
The image must either be free of copyright restrictions or belong to you. Read more.
Plagiarism of images or text is absolutely unacceptable. Read more.
All cases submitted should be completely de-identified, in such a way as to leave no way of identifying an individual patient. Ideally, images should be devoid of ALL text overlay. This not only removes visual clutter but ensures that no information is inadvertently included. Most PACS allow you to export without text overlay.
Information which should NOT appear in any of the cases includes (based on HIPAA guidelines), but is not limited to:
- date of birth
- address, including full or partial postal code
- telephone or fax numbers or other contact information
- e-mail addresses
- unique identifying numbers (e.g. UR, MRN, HID, etc)
- vehicle identifiers
- medical device identifiers (e.g. serial numbers)
- web or internet protocol addresses containing any link to the patient
- biometric data
- facial photograph or comparable image
- names of relatives
Additionally, if a case is for one reason or another unique in a way that could lead to the identification of an individual then it should also not be uploaded. This means that a case which has been, for example, featured in the media is usually inappropriate.
The vast majority of radiology departments now have PACS and with that, the ability to export images to disc. It is unlikely that taking a photograph of a screen will provide a case that is helpful for the illustration of a sign or disease process.
Since our aim is to create the best radiology resource of the internet, we need good quality images. Moreover, if the case includes cross-sectional imaging or DSA, the images should be uploaded as a stack. Ideally, these should be between 20-70 images per stack and excessive images (e.g. over 100) are generally discouraged; they take far too long to load and thus make your case difficult to access by most folk with lower speed connections.
You consider using one of our third-party uploaders. These make uploading cases a breeze.
- resolution: no specific figure is set. The minimum is the native resolution of the modality (e.g. 512x512px for CT, 128 to 1024px for MRI (depending on sequence / scan parameters), >1024px for plain films, etc).
- contrast/brightness: the abnormality should be visible easily and the image displayed with optimum windowing and levels (remember that these are not DICOM and therefore should be optimized at the time of capture).
Images should also be cropped appropriately so that there is no extraneous amount of blank space around them - read more.
Text overlays and watermarks
Obviously, images should include zero identifying information (see above). The best way to do this is to have zero text overlays.
The benefits of having no text overlays whatsoever are manyfold:
- ensures no patient information
- ensures that no one interprets text as patient information (even if it isn't)
- makes the image look great
- makes it easier to use the image in presentations
If you want to include an arrow or annotation, please upload a separate 'annotated' version of a single image demonstrating an ROI measurement is best.
For clinical images, no watermarks, logos or other graphics should be on the image. This includes:
- your name or logo or signature
- your institution's logo or name
- the brand of your PACS or viewer or image processor
- watermarks over the image of any kind
There are a couple of exceptions:
- manufacturer logos are sometimes burnt into the images (particularly ultrasound); if possible remove these also
- your signature and creative commons logo on illustrations (see medical illustrations and diagrams for more information)
Cases need to have a diagnosis, and each case has a diagnostic certainty slider at the top of the case to help you. This may be radiological (supracondylar fracture) or histological (eosinophilic granuloma). It is important that if the diagnosis is histological, appropriate mention is made in the discussion. Ideally, you should include the histology report (and even better - images of the slides).
If the diagnosis is a presumptive diagnosis pending histological confirmation, this should be clarified in the text.
In some cases, where no laboratory or pathological test is definitive, the diagnosis is presumptive. In such cases, you need to convince the readers/editors why your diagnosis is correct.
Just because you know what the image or stack of images shows, it does not mean that everybody does! So, add a short description of the study. It does not need to be a complete report but should highlight the salient features.
These are not mandatory, but can turn a good case into a great case!
The discussion portion of the case isn't mandatory but can be helpful for more complex cases. It should be limited to this case and delegate discussion of the condition in general to the "related article" section (see below). You can link some keywords to articles in Radiopaedia (read more about linking).
At the end of each case please add relevant related Radiopaedia articles (minimum of one, but usually two or three).