Radiopaedia Blog

Recently, following a case of a child who swallowed a pendant of a popular cartoon character (case here), a small flurry of media attention has been focused on Radiopaedia.org. Although some articles, including this one from ABC news in Australia, took the time to understand what we are about, many merely latched on to the more whimsical side of radiology.

Yes it is true we have a modest collection of rectal foreign bodies. Yes, they are a source of amusement; I would be lying if I didn't admit that I find the idea of someone presenting to the emergency department with an aubergine tucked up where the sun-don't-shine just a little bit funny.

These cases, however, represent a tiny fraction of our collection (~0.1% of cases and ~0.001% of images on the site; most of our 17000+ cases are detailed with multiple modalities and each comprise hundreds of images, description of the findings, and associated test results, e.g. biopsy results etc...). They are, therefore, at most a footnote to what we are about: building the most comprehensive, reliable and accessible radiology resource ever, to enable clinicians from every corner of the world to better diagnose and treat their patients. 

We believe that by sharing our collective knowledge, without imposing restrictions based on one's personal, institutional or geographical wealth, we can do a great deal of good. Over the past almost 10 years, we have made great progress, with our site now visited by around 2 million individuals every month, from every country on earth. The vast majority of these visits, well over 99%, are to non-sensational, detailed technical pages covering almost every topic in radiology. Many of the conditions we discuss are common but we also have many examples of very rare conditions; the sort of condition a radiologist may only ever see one example of during their career. By providing a platform where we can share such cases with our colleagues, we help each other make the right diagnosis, and in turn help our clinicians treat these patients better. 

So while we appreciate media attention, it would be great to not lose track of the real goal of Radiopaedia.org.

Frank

Dr Frank Gaillard is a neuroradiologist at the Royal Melbourne Hospital, Melbourne, Australia, and is the Founder and Editor of Radiopaedia.org.  NB: Opinions expressed are those of the author alone, and are not necessarily those of his employer. In this case they are very much the opinons of Radiopaedia.org.

Most of us know that intracranial epidermoid cysts demonstrate restricted diffusion, and we look for high signal on DWI to make the diagnosis. What I have repeatedly encountered, however, is confusion when the ADC maps are reviewed and the cyst content has signal similar to white matter, rather than being as black as an acute ischemic stroke. This leads many to start doubting the diagnosis, despite all imaging being characteristic. So in this quick post I will try and address the question "what do we mean when we say that epidermoid cysts demonstrate restricted diffusion", and to more generally address the use of the term "restricted diffusion". 

When we say "epidermoids demonstrate restricted diffusion" what we mean is "compared to arachnoid cysts"

The problem arises from the fact that we grow up being exposed to the concept of restricted diffusion in the context of ischemic stroke. We are used to saying "restricted diffusion is present consistent with an acute infarct". What we forget is that actually what we meaning is "this bit of brain restricts diffusion of water more than it should, were it to be normal". Because of this we think of diffusion restriction dichotomously (ie present or absent) rather than as a measure of a physical parameter, similar to say density on CT. So, when we say "epidermoids demonstrate restricted diffusion" what we mean is "compared to arachnoid cysts" which is usually their differential. The presence of a degree of restricted diffusion categorically tells you that this very high T2 signal lesion with low T1 and at least partial suppression on FLAIR does not contain CSF. In reality epidermoid cysts only moderately restrict, roughly similar to white matter. Most of the high DWI signal is due to T2 shine-through. 

Case 1

CASE 1: Typical right CPA epidermoid cyst, demonstrating vivid high DWI signal, and ADC value of  839 x 10^-6 mm²/s (see whole case)

Become familiar with normal ADC values and routinely measure them and include them in reports.

Normal values

I have been trying to encourage my registrars to become familiar with normal ADC values and routinely measure them and include them in reports. You should become as familiar with these values as you are with CT Hounsfield units (HU). So below are a few typical values (see ADC article for references; all values in x 10^-6 mm²/s) followed by case examples: 

• white matter: 670 - 800
• cortical grey matter:  800 - 1000
• deep grey matter: 700 - 850 
• CSF: 3000 - 3400

Case 2

Case 2: Normal posterior fossa, cerebellar white matter ADC of 766 x 10^-6 mm²/s and CSF of 3498 x 10^-6 mm²/s. 

Case 3

Case 3: Typical middle cranial fossa arachnoid cyst with ADC values of CSF 3136 x 10^-6 mm²/s (see whole case)

Case 4

Case 4: Middle cranial fossa epidermoid cyst with ADC values of 720 x 10^-6 mm²/s, similar to adjacent white matter  (see whole case) 

If you compare the ADC values from an epidermoid cyst of say 700 to 800 x 10^-6 mm²/s, to those obtained from acute infarcts or cerebral abscesses then you will note that the ADC values in both of those situations are much much lower (more restricted). 

Case 5

Case 5: Two day old infarct, with ADC values of 344 x x 10^-6 mm²/s.

Case 6

Case 6: Cerebral abscess  with ADC values of the pus of 500 x 10^-6 mm²/s (see whole case)

Take home message

So what is the take home message? Diffusion restriction is a continuous scale ranging from approximately 3400 for CSF all the way down to 0, and to say merely "restricted diffusion is present" is as meaningless as saying in a CT report that "density is present". Get used to the values you expect to encounter and remember that invariably when you are talking about the presence of restricted diffusion it is relative to some other tissue. As you become more comfortable with these values, you will find that you start using them in many other scenarios, such as predicting grade of gliomas and assessing treatment response in high grade tumors. 

Frank

Dr Frank Gaillard is a neuroradiologist at the Royal Melbourne Hospital, Melbourne, Australia, and is the Founder and Editor of Radiopaedia.org.  NB: Opinions expressed are those of the author alone, and are not those of his employer, or of Radiopaedia.org.

We have had a great year of case uploads - thank you to everyone who has contributed. Led by managing editor Dr Andrew Dixon, many of these cases were shared via social media and here we countdown the top 3 by views. 

Pneumothorax - bilateral contributed by Dr Hani Salam. A classic teaching case. 

Teapot syndrome with grade 4 acrofemoral synostosis contributed by Dr Matt Skalski. Another great April Fools' case that got a few people going! 

Intraperitoneal zucchini contributed by Dr Frank Gaillard. Well then... not too much to say here...

Dr Henry Knipe is a radiology registrar at The Royal Melbourne Hospital in Australia, and is managing editor responsible for content development at Radiopaedia.org. Twitter: @DrHenryK.

NB: Opinions expressed are those of the author alone, and are not those of his employer nor of Radiopaedia.org. 

Radiopaedia.org has always had as one of its core goals, to collect a huge library of cases and to make them readily available to others for teaching the art and science of radiology. To that end, all the cases contributed by 1000s of users are made available through a Creative Commons license. What does this mean? Well, it's really simple: as long as you abide by the rules of our license, you don't need to ask permission to use the images for free.

The license conditions are really simple: 

• your use must be non-commercial
• you must attribute the case to the contributing author
• you cannot slap your own copyright or different license on the image

Simple right? Well you would think so, but recently I traveled greater than twenty hours from my home town of Melbourne to a well known North American university hospital - you know who you are - for a moderately expensive (>$1000) neuroimaging seminar, and to my amazement I recognized about 10 of my cases, all unattributed. Not only that, but the handout for the lectures were all copyrighted. So all three creative commons license rules were violated. 

I experienced a number of different emotions in close succession: pleasure at seeing my cases, annoyance at not having them attributed to me, irritation at the fact that this institution was charging folk to see my cases which are free on Radiopaedia.org, and that they were using these cases to build their brand - "wow, they have great cases at institution X".

My initial knee-jerk response was "how can I stop this from happening?". When I had settled down and taken a breath I realized that as poor form as this is (you really should be ashamed of yourself) it is no reason to make it harder to use our images. If we did that, we would be undermining everything Radiopaedia.org stands for. It didn't actually cost me anything and 200 or so radiologists from around the world understand neuroradiology a little bit better, at least in part because of my images. That's a good thing, despite the circumstances. 

So today I want to introduce you to a little feature that makes it even easier to use our images correctly.

Icon by Thomas Le Bas * 

This little download button, located in the top left corner of images, will help you download a copy of the selected image for you to use in your presentation or handout or video or whatever. We have made it super easy for you to do the right thing, including cut-and-paste attribution and links. 

And remember, if you need to use an image outside of the creative commons license, just write to us and we and the contributing author will work something out. 

Thank you all, and keep sharing. Together we are making a real difference in how radiology is taught. 

Frank

* This icon by Thomas Le Bas is from the Noun Project (icon here). We have purchased the rights to use this icon on our website, but I include attribution to Thomas in this blog, not because we have to but just to show you that attribution is easy and gives you warm fuzzy feelings of knowing that you are giving credit where credit is due :) 

Dr Frank Gaillard is a neuroradiologist at the Royal Melbourne Hospital, Melbourne, Australia, and is the Founder and Editor of Radiopaedia.org.  NB: Opinions expressed are those of the author alone, and are not those of his employer, or of Radiopaedia.org.

Pareidolia is part of the art and culture of radiology. It is the ability or tendency to see faces, animals or objects when given vague stimuli. The classic example is seeing shapes in clouds. It is a subset of the more generic apophenia, which is merely seeing patterns in random data. 

Here is a lovely example featuring both a black butterfly and a white butterfly. 

Black and white butterflies (ventricles and meningioma)

Being on the look out for such little gems adds an unexpected delight to even the most mundane of studies. We would love you to contribute some examples, and share with us these hidden images. 

Dr Frank Gaillard is a neuroradiologist at the Royal Melbourne Hospital, Melbourne, Australia, and is the Founder and Editor of Radiopaedia.org.  NB: Opinions expressed are those of the author alone, and are not those of his employer, or of Radiopaedia.org.